AI model prices and specs
Every model available in Blend, with context window and price per million tokens. This table is generated from Blend’s model registry, which syncs automatically as providers ship new models.
Model data updated: 2026-10-04
| Model | Provider | Context | Input | Output | Images |
|---|---|---|---|---|---|
| Claude Fable 5 | Anthropic | 128K | $3 | $15 | Yes |
| Claude Haiku 4.5 | Anthropic | 200K | $0.80 | $4 | Yes |
| Claude Opus 5 | Anthropic | 128K | $15 | $75 | Yes |
| Claude Sonnet 5 | Anthropic | 128K | $3 | $15 | Yes |
| Command R | Cohere | 128K | $0.15 | $0.60 | — |
| Command R+ | Cohere | 128K | $2.50 | $10 | — |
| Gemini 2.5 Flash | 1.0M | $0.15 | $0.60 | Yes | |
| Gemini 2.5 Pro | 1.0M | $1.25 | $10 | Yes | |
| Gemini 3.5 Flash Lite | 1.0M | $0.10 | $0.40 | Yes | |
| Gemini 3.8 Flash | 1.0M | — | — | Yes | |
| Gemini Pro | 1.0M | $0.10 | $0.40 | Yes | |
| Gemma 4 26B IT | 262K | $0.10 | $0.40 | Yes | |
| Gemma 4 31B IT | 262K | $0.10 | $0.40 | Yes | |
| GPT-OSS 120B | Groq | 131K | $0.15 | $0.60 | — |
| GPT-OSS 20B | Groq | 131K | $0.07 | $0.30 | — |
| Qwen 3.6 27B | Groq | 131K | $0.60 | $3 | — |
| Codestral | Mistral | 256K | $0.30 | $0.90 | — |
| Mistral Large | Mistral | 128K | $2 | $6 | — |
| GPT-5.2 | OpenAI | 128K | $0.50 | $1.50 | Yes |
| GPT-5.4 mini | OpenAI | 128K | $0.50 | $1.50 | Yes |
| GPT-5.4 nano | OpenAI | 128K | $0.50 | $1.50 | Yes |
| GPT-5.6 Terra | OpenAI | 128K | $0.50 | $1.50 | Yes |
Input / Output: per 1M tokens. Prices are the providers’ list prices in USD per million tokens and can change at any time. Use them for comparison, not billing.
Frequently asked
ChatGPT — OpenAI
The most widely used general-purpose assistant, with the broadest ecosystem of tools and integrations.
Claude — Anthropic
Favoured for long documents, careful reasoning and natural writing, with very large context windows.
Gemini — Google
Strong price-to-performance with fast responses and generous free access — the model behind Blend’s free trial.
DeepSeek — DeepSeek
An open-weight family known for strong reasoning and coding at a very low price point.
Groq — Groq
Open models served on Groq hardware — the fastest first-token latency available in Blend.
Mistral — Mistral AI
European models with a strong efficiency focus and good multilingual coverage.
Blend — one question, multiple AIs, better answers
Related comparisons
- ChatGPT vs Claude: which should you use?
- ChatGPT vs Gemini: which should you use?
- Claude vs Gemini: which should you use?
- ChatGPT vs Groq: which should you use?
- The best AI for coding
- The best AI for writing
- The best AI for translation
Model capabilities and prices change often. Figures come from Blend’s automatically synced registry and are shown for comparison. All model prices · Terms · Privacy Policy · Refund