Cohere develops enterprise AI models for generation, embeddings, reranking, translation, document parsing and speech recognition. Its models support RAG, search, agents and multilingual applications. Shared API pricing, evaluation access and dedicated deployment terms are listed separately.
Usage-based API; see model prices.
Listed models · 31
Distinct model versions are listed once; aliases are not additional models. Prices below are for API usage.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command A+
Official model ID: command-a-plus-05-2026
Model developer: Cohere · Service provider: Self-hosted
Model family: Command · Model catalog only; see service-channel quotes
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
Model context: 128k · Official rounded display; exact token count unverified
Maximum output: 64k · Official rounded display
Input / output modalities: text, image → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Contact Cohere for production access.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command A
Official model ID: command-a-03-2025
Model developer: Cohere · Service provider: Self-hosted
Model family: Command · Model catalog only; see service-channel quotes
The official launch article still publishes USD 2.50 input / 10.00 output per 1M tokens; the current pricing-page cards omit Command A. Treat this as official-announcement pricing requiring reconfirmation, not a newly advertised card price.
Model context: 256k · Official rounded display; exact token count unverified
Maximum output: 8k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
The official launch article still publishes USD 2.50 input / 10.00 output per 1M tokens; the current pricing-page cards omit Command A. Treat this as official-announcement pricing requiring reconfirmation, not a newly advertised card price.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command R7B
Official model ID: command-r7b-12-2024
Model developer: Cohere · Service provider: Cohere
Model family: Command · Currency: USD
Rates per million tokens| Price dimension | Current rate |
|---|
| Input | USD 0.0375 |
|---|
| Output | USD 0.15 |
|---|
| Cached read | Unverified |
|---|
Documented API context: 128k · Official rounded display; exact token count unverified
Maximum output: 4k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command A Translate
Official model ID: command-a-translate-08-2025
Model developer: Cohere · Service provider: Self-hosted
Model family: Command · Model catalog only; see service-channel quotes
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
Model context: 8K · Official rounded display; exact token count unverified
Maximum output: 8k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Contact Cohere for production access.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command A Reasoning
Official model ID: command-a-reasoning-08-2025
Model developer: Cohere · Service provider: Self-hosted
Model family: Command · Model catalog only; see service-channel quotes
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
Model context: 256k · Official rounded display; exact token count unverified
Maximum output: 32k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Contact Cohere for production access.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command A Vision
Official model ID: command-a-vision-07-2025
Model developer: Cohere · Service provider: Self-hosted
Model family: Command · Model catalog only; see service-channel quotes
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
Model context: 128K · Official rounded display; exact token count unverified
Maximum output: 8K · Official rounded display
Input / output modalities: text, image → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Contact Cohere for production access.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Production keys use trial limits for this newer model; contact Cohere for production access. No standard shared-API production token price was verified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command R (08-2024)
Official model ID: command-r-08-2024
Model developer: Cohere · Service provider: Cohere
Model family: Command · Currency: USD
Rates per million tokens| Price dimension | Current rate |
|---|
| Input | USD 0.15 |
|---|
| Output | USD 0.6 |
|---|
| Cached read | Unverified |
|---|
Documented API context: 128k · Official rounded display; exact token count unverified
Maximum output: 4k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Command R+ (08-2024)
Official model ID: command-r-plus-08-2024
Model developer: Cohere · Service provider: Cohere
Model family: Command · Currency: USD
Rates per million tokens| Price dimension | Current rate |
|---|
| Input | USD 2.5 |
|---|
| Output | USD 10 |
|---|
| Cached read | Unverified |
|---|
Documented API context: 128k · Official rounded display; exact token count unverified
Maximum output: 4k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
North Small Translate 1.0
Official model ID: north-small-translate-1-0
Model developer: Cohere · Service provider: Self-hosted
Model family: North · Model catalog only; see service-channel quotes
The official catalog describes hosted API access as evaluation. Free API access is not a production-free offering.
Model context: 16k · Official rounded display; exact token count unverified
Maximum output: 16k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Open weights are CC BY-NC 4.0 for non-commercial use. Confirm commercial terms with Cohere.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
The official catalog describes hosted API access as evaluation. Free API access is not a production-free offering.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
North Mini Code 1.0
Official model ID: north-mini-code-1-0
Model developer: Cohere · Service provider: Self-hosted
Model family: North · Model catalog only; see service-channel quotes
The official catalog describes hosted API access as evaluation. Free API access is not a production-free offering.
Model context: 256k · Official rounded display; exact token count unverified
Maximum output: 64k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Production deployment is offered through Model Vault. Apache 2.0 applies to released weights, not free hosted inference.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
The official catalog describes hosted API access as evaluation. Free API access is not a production-free offering.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Embed 5 Pro
Official model ID: embed-v5.0-pro
Model developer: Cohere · Service provider: Cohere
Model family: Embed · Currency: USD
Rates per million tokens| Price dimension | Current rate |
|---|
| Input | USD 0.12 |
|---|
| Output | Unverified |
|---|
| Cached read | Unverified |
|---|
Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
Documented API context: 128k · Official rounded display; exact token count unverified
Input / output modalities: text, image, mixed_text_image → embedding
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Embed 5 Fast
Official model ID: embed-v5.0-fast
Model developer: Cohere · Service provider: Cohere
Model family: Embed · Currency: USD
Rates per million tokens| Price dimension | Current rate |
|---|
| Input | USD 0.08 |
|---|
| Output | Unverified |
|---|
| Cached read | Unverified |
|---|
Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
Documented API context: 128k · Official rounded display; exact token count unverified
Input / output modalities: text, image, mixed_text_image → embedding
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Embed 4
Official model ID: embed-v4.0
Model developer: Cohere · Service provider: Self-hosted
Model family: Embed · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
Model context: 128k · Official rounded display; exact token count unverified
Input / output modalities: text, image, mixed_text_image → embedding
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Embed 3 English
Official model ID: embed-english-v3.0
Model developer: Cohere · Service provider: Self-hosted
Model family: Embed · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
Model context: 512 tokens
Input / output modalities: text, image → embedding
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Embed 3 English Light
Official model ID: embed-english-light-v3.0
Model developer: Cohere · Service provider: Self-hosted
Model family: Embed · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
Model context: 512 tokens
Input / output modalities: text, image → embedding
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Embed 3 Multilingual
Official model ID: embed-multilingual-v3.0
Model developer: Cohere · Service provider: Self-hosted
Model family: Embed · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
Model context: 512 tokens
Input / output modalities: text, image → embedding
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Embed 3 Multilingual Light
Official model ID: embed-multilingual-light-v3.0
Model developer: Cohere · Service provider: Self-hosted
Model family: Embed · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
Model context: 512 tokens
Input / output modalities: text, image → embedding
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Embedding has no generated-output token charge. The published token rate does not establish a separate image price, image-to-token conversion, or batch discount; these remain unverified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Rerank 4 Pro
Official model ID: rerank-v4.0-pro
Model developer: Cohere · Service provider: Cohere
Model family: Rerank · Currency: USD
- search: Unverified; see official rules
Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
Documented API context: 32,768 tokens
Input / output modalities: text → relevance_scores
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Rerank 4 Fast
Official model ID: rerank-v4.0-fast
Model developer: Cohere · Service provider: Cohere
Model family: Rerank · Currency: USD
- search: Unverified; see official rules
Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
Documented API context: 32,768 tokens
Input / output modalities: text → relevance_scores
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Rerank 3.5
Official model ID: rerank-v3.5
Model developer: Cohere · Service provider: Self-hosted
Model family: Rerank · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
Model context: 4,096 tokens
Input / output modalities: text → relevance_scores
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Rerank 3 English
Official model ID: rerank-english-v3.0
Model developer: Cohere · Service provider: Self-hosted
Model family: Rerank · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
Model context: 4,096 tokens
Input / output modalities: text → relevance_scores
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Rerank 3 Multilingual
Official model ID: rerank-multilingual-v3.0
Model developer: Cohere · Service provider: Self-hosted
Model family: Rerank · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
Model context: 4,096 tokens
Input / output modalities: text → relevance_scores
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero. Billed per search, not per input/output token. Pricing defines a search as one query with up to 100 documents; documents exceeding 500 tokens including query length are split into billable chunks. Use API billed_units.search_units for the exact charge.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Parse 5
Official model ID: parse-v5.0
Model developer: Cohere · Service provider: Cohere
Model family: Parse · Currency: USD
- page: Unverified; see official rules
Shared API price is USD 1.50 per 1,000 pages. Model Vault pricing is separate. Returns Markdown; structured JSON output is not supported.
Documented API context: 8,192 tokens
Input / output modalities: document, image → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Shared API price is USD 1.50 per 1,000 pages. Model Vault pricing is separate. Returns Markdown; structured JSON output is not supported.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Cohere Transcribe
Official model ID: cohere-transcribe-03-2026
Model developer: Cohere · Service provider: Self-hosted
Model family: Transcribe · Model catalog only; see service-channel quotes
Hosted API is for free rate-limited experimentation (5 requests/minute). Production uses dedicated Model Vault billed per instance-hour; no per-audio-minute or per-token rate was verified.
Input / output modalities: audio → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Production uses a dedicated Model Vault; contact Cohere for access.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Hosted API is for free rate-limited experimentation (5 requests/minute). Production uses dedicated Model Vault billed per instance-hour; no per-audio-minute or per-token rate was verified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Cohere Transcribe Arabic
Official model ID: cohere-transcribe-arabic-07-2026
Model developer: Cohere · Service provider: Self-hosted
Model family: Transcribe · Model catalog only; see service-channel quotes
Hosted API is for free rate-limited experimentation (5 requests/minute). Production uses dedicated Model Vault billed per instance-hour; no per-audio-minute or per-token rate was verified.
Input / output modalities: audio → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Production uses a dedicated Model Vault; contact Cohere for access.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
Hosted API is for free rate-limited experimentation (5 requests/minute). Production uses dedicated Model Vault billed per instance-hour; no per-audio-minute or per-token rate was verified.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Tiny Aya Global
Official model ID: tiny-aya-global
Model developer: Cohere · Service provider: Self-hosted
Model family: Aya · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
Model context: 8k · Official rounded display; exact token count unverified
Maximum output: 8k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Tiny Aya Earth
Official model ID: tiny-aya-earth
Model developer: Cohere · Service provider: Self-hosted
Model family: Aya · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
Model context: 8k · Official rounded display; exact token count unverified
Maximum output: 8k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Tiny Aya Fire
Official model ID: tiny-aya-fire
Model developer: Cohere · Service provider: Self-hosted
Model family: Aya · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
Model context: 8k · Official rounded display; exact token count unverified
Maximum output: 8k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Tiny Aya Water
Official model ID: tiny-aya-water
Model developer: Cohere · Service provider: Self-hosted
Model family: Aya · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
Model context: 8k · Official rounded display; exact token count unverified
Maximum output: 8k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Aya Expanse 32B
Official model ID: c4ai-aya-expanse-32b
Model developer: Cohere · Service provider: Cohere
Model family: Aya · Currency: USD
Rates per million tokens| Price dimension | Current rate |
|---|
| Input | USD 0.5 |
|---|
| Output | USD 1.5 |
|---|
| Cached read | Unverified |
|---|
Documented API context: 128k · Official rounded display; exact token count unverified
Maximum output: 4k · Official rounded display
Input / output modalities: text → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
API models and prices · 1
Usage-based billing: text rates are usually per million tokens; other units and tiers are stated per entry. Compare only the same currency, unit and billing conditions. Unverified cache pricing does not mean free.
Aya Vision 32B
Official model ID: c4ai-aya-vision-32b
Model developer: Cohere · Service provider: Self-hosted
Model family: Aya · Model catalog only; see service-channel quotes
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
Model context: 16k · Official rounded display; exact token count unverified
Maximum output: 4k · Official rounded display
Input / output modalities: text, image → text
Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised. Trial API keys are free for evaluation, limited to 1,000 API calls per month, and may not be used for production or commercial purposes. No monetary credit or unlimited allowance is promised.
Official source · Last verified: 2026-10-15
Capabilities and usage limits
No current numeric shared-API price was verified in the current English pricing page or linked model documentation. Missing prices are not zero.
