AI Rate CardAI API prices compared: ChatGPT (GPT-6), Claude, Gemini, DeepSeek and Qwen
Updated Nov 3, 2026
AI APIs are billed by tokens: what you send (input) and what the model answers (output), priced separately. A ChatGPT Plus or Claude Pro subscription is access to the app; through the API the same models are sold per token, and dozens of providers resell them. Below: the typical price (median across providers) of the five main model families, and live tables rebuilt from today's prices on every page view.
Five families in one table
Price per 1M tokens, input / output, 2 November 2026 collection, median across providers:
| Developer | Flagship | Mid-range | Cheap |
|---|---|---|---|
| OpenAI (ChatGPT) | GPT-6 Astra $10 / $50 | GPT-6.1 Sol $2 / $10 | GPT-6 Luna $0.10 / $0.50 |
| Anthropic (Claude) | Claude Opus 5.5 $4 / $20 | Claude Sonnet 5.5 $2 / $10 | Claude Haiku 5.5 $0.10 / $0.50 |
| Google (Gemini) | Gemini 4 Argon $2 / $10 | Gemini 3.8 Flash $0.75 / $3.75 | Gemini 3.5 Flash-Lite $0.30 / $2.50 |
| DeepSeek | DeepSeek V4 Pro $1.53 / $3.30 | DeepSeek V4.1 Flash $0.30 / $1.20 | DeepSeek V4 Flash $0.14 / $0.30 |
| Alibaba (Qwen) | Qwen3.8 Max $2 / $6 | Qwen3.8 27B $0.40 / $2.50 | Qwen3.8 Flash $0.15 / $0.47 |
- Cheapest: GPT-6 Luna and Claude Haiku 5.5 ($0.10 / $0.50), Qwen3.8 Flash and DeepSeek V4 Flash.
- Strong models at a mid price: GPT-6.1 Sol, Claude Sonnet 5.5 and Gemini 4 Argon all cost $2 / $10; Qwen3.8 Max is cheaper on output ($6).
- Most expensive: GPT-6 Astra ($10 / $50) and Claude Opus 5.5.
- Open-weight models (DeepSeek, Qwen) have the widest spread between providers - compare the tables below.
ChatGPT (OpenAI): GPT-6.1 Sol at every provider
| Provider | Model | Input, $ | Output, $ |
|---|---|---|---|
| openrouter | OpenAI: GPT-6.1 Sol (batch) | $1.00 | $5.00 |
| requesty | GPT-6.1 Sol | $1.00 | $5.00 |
| vercel | GPT-6.1 Sol (Batch) | $1.00 | $5.00 |
| cometapi | GPT-6.1 Sol | $1.60 | $8.00 |
| iteracompute | GPT-6.1 Sol | $1.60 | $8.00 |
| openai | GPT-6.1 Sol | $2.00 | $10.00 |
| 302ai | GPT-6.1 Sol | $2.00 | $10.00 |
| abacus | GPT-6.1 Sol | $2.00 | $10.00 |
| aihubmix | GPT 6.1 Sol | $2.00 | $10.00 |
| aiml | GPT-6.1 Sol | $2.00 | $10.00 |
Live data: this table is built from today's prices.
Claude: Sonnet 5.5 at every provider
| Provider | Model | Input, $ | Output, $ |
|---|---|---|---|
| iteracompute | Claude Sonnet 5.5 | $1.00 | $5.00 |
| openrouter | Anthropic: Claude Sonnet 5.5 (batch) | $1.00 | $5.00 |
| vercel | Claude Sonnet 5.5 (Batch) | $1.00 | $5.00 |
| cometapi | Claude Sonnet 5.5 | $1.60 | $8.00 |
| anthropic | Claude Sonnet 5.5 | $2.00 | $10.00 |
| 302ai | Claude Sonnet 5.5 | $2.00 | $10.00 |
| abacus | Claude Sonnet 5.5 | $2.00 | $10.00 |
| aihubmix | Claude Sonnet 5.5 | $2.00 | $10.00 |
| aiml | Claude Sonnet 5.5 | $2.00 | $10.00 |
| amazon-bedrock | Claude Sonnet 5.5 | $2.00 | $10.00 |
Live data: this table is built from today's prices.
Gemini: 3.8 Flash at every provider
| Provider | Model | Input, $ | Output, $ |
|---|---|---|---|
| kenari | Gemini 3.8 Flash | $0.33 | $1.65 |
| openrouter | Google: Gemini 3.8 Flash (batch) | $0.38 | $1.88 |
| cometapi | Gemini 3.8 Flash | $0.60 | $3.00 |
| iteracompute | Gemini 3.8 Flash | $0.60 | $3.00 |
| edenai | Gemini 3.8 Flash | $0.75 | $3.75 |
| edenai | Gemini 3.8 Flash | $0.75 | $3.75 |
| Gemini 3.8 Flash | $0.75 | $3.75 | |
| 302ai | Gemini 3.8 Flash | $0.75 | $3.75 |
| abacus | Gemini 3.8 Flash | $0.75 | $3.75 |
| aihubmix | Gemini 3.8 Flash | $0.75 | $3.75 |
Live data: this table is built from today's prices.
DeepSeek: V4.1 Flash at every provider
| Provider | Model | Input, $ | Output, $ |
|---|---|---|---|
| nvidia | DeepSeek V4.1 Flash | Free | Free |
| opencode | DeepSeek V4.1 Flash | Free | Free |
| openrouter | DeepSeek: DeepSeek V4.1 Flash | $0.045 | $0.50 |
| openrouter | DeepSeek: DeepSeek V4.1 Flash (batch) | $0.05 | $0.15 |
| edenai | DeepSeek V4.1 Flash | $0.0975 | $0.39 |
| neuralwatt | DeepSeek V4.1 Flash Flex | $0.0975 | $0.39 |
| cometapi | DeepSeek V4.1 Flash | $0.12 | $0.48 |
| deepinfra | DeepSeek V4.1 Flash | $0.14 | $0.42 |
| fastrouter | DeepSeek: DeepSeek V4.1 Flash | $0.14 | $0.42 |
| iteracompute | DeepSeek V4.1 Flash | $0.14 | $0.55 |
Live data: this table is built from today's prices.
Qwen: 3.8 Max at every provider
| Provider | Model | Input, $ | Output, $ |
|---|---|---|---|
| edenai | Qwen3.8 Max | $1.30 | $3.90 |
| cometapi | Qwen3.8 Max | $1.60 | $4.80 |
| deepinfra | Qwen3.8 Max | $1.65 | $4.95 |
| edenai | Qwen3.8 Max | $1.65 | $4.95 |
| empiriolabs | Qwen3.8 Max | $1.65 | $4.95 |
| fastrouter | Qwen: Qwen3.8 Max | $1.65 | $4.95 |
| aihubmix | Qwen3.8 Max | $1.69 | $5.07 |
| ofox | Qwen3.8 Max | $1.71 | $5.14 |
| trustedrouter | Qwen3.8 Max | $1.74 | $5.22 |
| moark | Qwen3.8 Max | $1.80 | $5.40 |
Live data: this table is built from today's prices.
Each family in detail
ChatGPT/GPT, Claude, Gemini, DeepSeek, Qwen. Estimate a month for your workload with the calculator; the cheapest models overall: the cheapest LLM APIs.
All prices are in USD per 1 million tokens, without taxes, as the providers publish them; each row links to its card with the source and the date it was checked.