What 37 AI models cost, read from each maker’s own price list
The list price of every model a small business is likely to weigh, from 11 labs, and what three everyday jobs cost on each: a support reply, a document summary and an agent task.
Prices in US dollars per million tokens, at the standard rate. No reseller prices.
- Models
- 37
- Makers
- 11
- Open weights
- 15
- No list price
- 4
3,000 tokens in and 400 out each, at list price.
We read the list price of 37 AI models from 11 makers on their own pricing pages on 22 September 2026. On 1,000 support replies, the cheapest, Qwen3.8 Flash, costs $0.64 and the most expensive, GPT-6 Astra, $50.00: a gap of 78 times for the same job.
Flagships
Each lab’s most capable model, the one it positions for the hardest work.
| Model | Input | Output | Context | 1,000 replies |
|---|---|---|---|---|
| MiniMax M3MiniMax · open weights | $0.30 | $1.20 | 1M | $1.38 |
| Muse Spark 1.3Meta | $1.25 | $4.25 | 1.05M | $5.45 |
| DeepSeek V4 ProDeepSeek · open weights | $1.32 | $3.96 | 1M | $5.54 |
| GLM-5.3Z.ai · open weights | $1.40 | $4.40 | 1M | $5.96 |
| Mistral Medium 3.5Mistral · open weights | $1.50 | $7.50 | 256K | $7.50 |
| Grok 4.7xAI | $2.00 | $6.00 | 500K | $8.40 |
| Qwen3.8 MaxAlibaba (Qwen) | $2.00 | $6.00 | 1M | $8.40 |
| Gemini 3.1 Pro (preview)Google | $2.00 | $12.00 | 1.05M | $10.80 |
| Kimi K3Moonshot AI · open weights | $3.00 | $15.00 | 1.05M | $15.00 |
| GPT-5.6 SolOpenAI | $4.00 | $20.00 | 1.05M | $20.00 |
| Claude Opus 5Anthropic | $5.00 | $25.00 | 1M | $25.00 |
| GPT-5.5OpenAI | $5.00 | $30.00 | 1.05M | $27.00 |
| Claude Fable 5.1Anthropic | $10.00 | $50.00 | 1M | $50.00 |
| GPT-6 AstraOpenAI | $10.00 | $50.00 | 1.05M | $50.00 |
Workhorses
The middle of each lab’s range: the model most products run on day to day.
| Model | Input | Output | Context | 1,000 replies |
|---|---|---|---|---|
| Qwen3.7 PlusAlibaba (Qwen) | $0.40 | $1.60 | 1M | $1.84 |
| Gemini 3.8 FlashGoogle | $0.75 | $3.75 | 1.05M | $3.75 |
| Kimi K2.6Moonshot AI · open weights | $0.95 | $4.00 | 262K | $4.45 |
| Claude Sonnet 5Anthropic | $2.00 | $10.00 | 1M | $10.00 |
| GPT-5.6 TerraOpenAI | $2.00 | $12.00 | 1.05M | $10.80 |
| Claude Sonnet 4.6Anthropic | $3.00 | $15.00 | 1M | $15.00 |
| gpt-oss-120bOpenAI · open weights | No list price | · | 131K | · |
| Llama 4 MaverickMeta · open weights | No list price | · | 1M | · |
Small and fast
The cheapest and quickest model each lab sells, for high volumes and simple steps.
| Model | Input | Output | Context | 1,000 replies |
|---|---|---|---|---|
| Qwen3.8 FlashAlibaba (Qwen) | $0.15 | $0.47 | 1M | $0.64 |
| GLM-5.3-FlashZ.ai · open weights | $0.15 | $0.50 | 1M | $0.65 |
| Ministral 3 14BMistral · open weights | $0.20 | $0.20 | 256K | $0.68 |
| Mistral Small 4Mistral · open weights | $0.15 | $0.60 | 256K | $0.69 |
| GPT-5.6 LunaOpenAI | $0.20 | $1.20 | 1.05M | $1.08 |
| GPT-5.4 nanoOpenAI | $0.20 | $1.25 | 400K | $1.10 |
| Gemini 3.1 Flash-LiteGoogle | $0.25 | $1.50 | 1.05M | $1.35 |
| DeepSeek V4.1 FlashDeepSeek · open weights | $0.30 | $1.20 | 1M | $1.38 |
| Gemini 3.5 Flash-LiteGoogle | $0.30 | $2.50 | 1.05M | $1.90 |
| Qwen3.8 27BAlibaba (Qwen) · open weights | $0.50 | $3.00 | 1M | $2.70 |
| GPT-5.4 miniOpenAI | $0.75 | $4.50 | 400K | $4.05 |
| Grok 4.3xAI | $1.25 | $2.50 | 1M | $4.75 |
| Claude Haiku 4.5Anthropic | $1.00 | $5.00 | 200K | $5.00 |
| Gemma 4 31BGoogle · open weights | No list price | · | 256K | · |
| Muse Glimmer 30BMeta · open weights | No list price | · | 128K | · |
The three jobs every cost is worked on
The token counts are ours, chosen to look like real work; the prices are the makers’. Every model page works all three.
- Support reply
- A 3,000-token prompt (the ticket, the thread so far and a policy excerpt) and a 400-token answer. 3,000 in, 400 out.
- Document summary
- A 40-page document of about 25,000 tokens in, a 1,000-token summary out. 25,000 in, 1,000 out.
- Agent task
- Ten steps that read 80,000 tokens between them, 70,000 of them context repeated from step to step and served from cache, and write 3,000. 80,000 in (70,000 cached), 3,000 out.
Tiers follow each maker’s own line-up, not a price band, so one lab’s flagship can cost less than another’s workhorse. 4 open-weight models have no list price because their makers sell no API for them; what they cost depends on the host.
Head to head
The 52 comparisons people search for, each with both prices and the three jobs costed on each.
- GPT-6 Astra vs Claude Fable 5.1
- GPT-6 Astra vs Gemini 3.1 Pro (preview)
- Claude Fable 5.1 vs Gemini 3.1 Pro (preview)
- GPT-6 Astra vs GPT-5.6 Sol
- Claude Fable 5.1 vs Claude Opus 5
- GPT-5.5 vs GPT-5.6 Sol
- Claude Opus 5 vs GPT-5.6 Sol
- Claude Opus 5 vs Gemini 3.1 Pro (preview)
- GPT-5.6 Sol vs Gemini 3.1 Pro (preview)
- Grok 4.7 vs GPT-5.6 Sol
- Grok 4.7 vs Claude Opus 5
- Grok 4.7 vs Gemini 3.1 Pro (preview)
- DeepSeek V4 Pro vs Claude Opus 5
- DeepSeek V4 Pro vs GPT-5.6 Sol
- Kimi K3 vs Claude Opus 5
- Qwen3.8 Max vs DeepSeek V4 Pro
- GLM-5.3 vs DeepSeek V4 Pro
- GLM-5.3 vs Kimi K3
- MiniMax M3 vs DeepSeek V4 Pro
- Muse Spark 1.3 vs GPT-5.6 Sol
- Muse Spark 1.3 vs Grok 4.7
- Mistral Medium 3.5 vs Claude Sonnet 5
- Claude Sonnet 5 vs GPT-5.6 Terra
- Claude Sonnet 5 vs Gemini 3.8 Flash
- GPT-5.6 Terra vs Gemini 3.8 Flash
- Claude Sonnet 5 vs Claude Sonnet 4.6
- Claude Sonnet 5 vs Claude Opus 5
- GPT-5.6 Sol vs GPT-5.6 Terra
- Qwen3.7 Plus vs Gemini 3.8 Flash
- Kimi K2.6 vs GPT-5.6 Terra
- Mistral Medium 3.5 vs Gemini 3.8 Flash
- Grok 4.3 vs Gemini 3.8 Flash
- Llama 4 Maverick vs gpt-oss-120b
- gpt-oss-120b vs DeepSeek V4.1 Flash
- GPT-5.6 Luna vs Gemini 3.5 Flash-Lite
- GPT-5.6 Luna vs Gemini 3.1 Flash-Lite
- Claude Haiku 4.5 vs GPT-5.6 Luna
- Claude Haiku 4.5 vs Gemini 3.8 Flash
- Claude Haiku 4.5 vs GPT-5.4 mini
- GPT-5.4 mini vs GPT-5.6 Luna
- GPT-5.4 nano vs GPT-5.6 Luna
- Gemini 3.5 Flash-Lite vs Gemini 3.1 Flash-Lite
- DeepSeek V4.1 Flash vs GPT-5.6 Luna
- DeepSeek V4.1 Flash vs Gemini 3.5 Flash-Lite
- Qwen3.8 Flash vs DeepSeek V4.1 Flash
- GLM-5.3-Flash vs Qwen3.8 Flash
- Mistral Small 4 vs GPT-5.6 Luna
- Grok 4.3 vs DeepSeek V4.1 Flash
- Ministral 3 14B vs Gemma 4 31B
- Gemma 4 31B vs Muse Glimmer 30B
- Muse Glimmer 30B vs Qwen3.8 27B
- Qwen3.8 27B vs Gemma 4 31B
Where the prices come from
Each price was read on the maker’s own pricing page or model documentation. These are the pages.
- Mistral
- docs.mistral.ai/models/mistral-medium-3-5-26-04 · docs.mistral.ai/models/mistral-small-4-0-26-03 · docs.mistral.ai/models/ministral-3-14b-25-12
- Alibaba (Qwen)
- alibabacloud.com/help/en/model-studio/model-pricing
- Moonshot AI
- platform.kimi.ai/docs/pricing/chat
Read on 22 September 2026. Each price is the standard rate for the shortest context band, with no batch or priority discount; long-context surcharges, batch rates and promotions are in each model’s small print. No price was taken from a reseller.
Pay a quarter of these prices
AIGROW API credit is metered at the makers’ list prices, like the ones above, and $25 buys $100 of usage. The credit page lists the models it covers.
Prices checked 22 September 2026