Price list
Every world, priced in the open
Each row shows what the upstream publishes and what Antimatter charges, side by side. The difference is the margin, and it is stated, not hidden in a rate.
+10%
The upstream's published price plus 10%. That is the whole margin, on every world.
The same for every account. No fee to add credit: what you send is what you are credited. It is not cheaper than buying direct and does not pretend to be: it is the upstream's price plus this, for one key, one wire and every world.
- 464
- 458
- 65
- Google: Gemini 3.6 Flash (batch)google/gemini-3.6-flash:batchvisiontoolsjson1M contextIn / 1M$0.4125up $0.375Out / 1M$2.063up $1.875Margin+10%1st tokenunmeasured
- Google: Gemini 3.7 Flash (batch)google/gemini-3.7-flash:batchvisiontoolsjson1M contextIn / 1M$0.4125up $0.375Out / 1M$2.063up $1.875Margin+10%1st tokenunmeasured
- Google: Gemini 3.8 Flash (batch)google/gemini-3.8-flash:batchvisiontoolsjson1M contextIn / 1M$0.4125up $0.375Out / 1M$2.063up $1.875Margin+10%1st tokenunmeasured
- Mistral: Mistral Large 3 2512mistralai/mistral-large-2512visiontoolsjson262k contextIn / 1M$0.55up $0.5Out / 1M$1.65up $1.50Margin+10%1st tokenunmeasured
- OpenAI: GPT-3.5 Turboopenai/gpt-3.5-turbotoolsjson16k contextIn / 1M$0.55up $0.5Out / 1M$1.65up $1.50Margin+10%1st tokenunmeasured
- Qwen: Qwen3 30B A3B Thinking 2507qwen/qwen3-30b-a3b-thinking-2507toolsjson82k contextIn / 1M$0.22up $0.2Out / 1M$2.64up $2.40Margin+10%1st tokenunmeasured
- Qwen: Qwen3 VL 30B A3B Thinkingqwen/qwen3-vl-30b-a3b-thinkingvisiontoolsjson262k contextIn / 1M$0.22up $0.2Out / 1M$2.64up $2.40Margin+10%1st tokenunmeasured
- Z.ai: GLM 4.6z-ai/glm-4.6toolsjson205k contextIn / 1M$0.473up $0.43Out / 1M$1.925up $1.75Margin+10%1st tokenunmeasured
- MoonshotAI: Kimi K2.6moonshotai/kimi-k2.6visiontoolsjson262k contextIn / 1M$0.4776up $0.4341Out / 1M$2.011up $1.828Margin+10%1st tokenunmeasured
- Qwen: Qwen3 235B A22Bqwen/qwen3-235b-a22btoolsjson131k contextIn / 1M$0.5005up $0.455Out / 1M$2.002up $1.82Margin+10%1st tokenunmeasured
- Mistral: Devstral 2 2512mistralai/devstral-2512toolsjson262k contextIn / 1M$0.44up $0.4Out / 1M$2.20up $2.00Margin+10%1st tokenunmeasured
- Mistral: Mistral Medium 3mistralai/mistral-medium-3visiontoolsjson131k contextIn / 1M$0.44up $0.4Out / 1M$2.20up $2.00Margin+10%1st tokenunmeasured
- Mistral: Mistral Medium 3.1mistralai/mistral-medium-3.1visiontoolsjson131k contextIn / 1M$0.44up $0.4Out / 1M$2.20up $2.00Margin+10%1st tokenunmeasured
- Z.ai: GLM 5.3 (batch)z-ai/glm-5.3:batchtoolsjson1M contextIn / 1M$0.495up $0.45Out / 1M$2.20up $2.00Margin+10%1st tokenunmeasured
- OpenAI: GPT-5.4 Mini (batch)openai/gpt-5.4-mini:batchvisiontoolsjson400k contextIn / 1M$0.4125up $0.375Out / 1M$2.475up $2.25Margin+10%1st tokenunmeasured
- Amazon: Nova 2 Liteamazon/nova-2-lite-v1visiontools1M contextIn / 1M$0.33up $0.3Out / 1M$2.75up $2.50Margin+10%1st tokenunmeasured
- Google: Gemini 2.5 Flashgoogle/gemini-2.5-flashvisiontoolsjson1M contextIn / 1M$0.33up $0.3Out / 1M$2.75up $2.50Margin+10%1st tokenunmeasured
- Google: Nano Banana (Gemini 2.5 Flash Image)google/gemini-2.5-flash-imagevisionjson33k contextIn / 1M$0.33up $0.3Out / 1M$2.75up $2.50Margin+10%1st tokenunmeasured
- Google: Gemini 3.5 Flash Litegoogle/gemini-3.5-flash-litevisiontoolsjson1M contextIn / 1M$0.33up $0.3Out / 1M$2.75up $2.50Margin+10%1st tokenunmeasured
- Qwen: Qwen2.5 VL 72B Instructqwen/qwen2.5-vl-72b-instructvisionjson128k contextIn / 1M$0.88up $0.8Out / 1M$1.10up $1.00Margin+10%1st tokenunmeasured
- Sao10K: Llama 3.1 Euryale 70B v2.2sao10k/l3.1-euryale-70btoolsjson131k contextIn / 1M$0.935up $0.85Out / 1M$0.935up $0.85Margin+10%1st tokenunmeasured
- AionLabs: Aion-3.0-Miniaion-labs/aion-3.0-minitoolsjson131k contextIn / 1M$0.77up $0.7Out / 1M$1.54up $1.40Margin+10%1st tokenunmeasured
- AionLabs: Aion 3.5 Miniaion-labs/aion-3.5-minitoolsjson262k contextIn / 1M$0.77up $0.7Out / 1M$1.54up $1.40Margin+10%1st tokenunmeasured
- MoonshotAI: Kimi K2.5moonshotai/kimi-k2.5visiontoolsjson262k contextIn / 1M$0.495up $0.45Out / 1M$2.475up $2.25Margin+10%1st tokenunmeasured
- Morph: Morph V3 Fastmorph/morph-v3-fast82k contextIn / 1M$0.88up $0.8Out / 1M$1.32up $1.20Margin+10%1st tokenunmeasured
- Z.ai: GLM 4.5Vz-ai/glm-4.5vvisiontoolsjson66k contextIn / 1M$0.66up $0.6Out / 1M$1.98up $1.80Margin+10%1st tokenunmeasured
- DeepSeek: R1 0528deepseek/deepseek-r1-0528toolsjson164k contextIn / 1M$0.55up $0.5Out / 1M$2.365up $2.15Margin+10%1st tokenunmeasured
- NVIDIA: Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55btoolsjson262k contextIn / 1M$0.55up $0.5Out / 1M$2.42up $2.20Margin+10%1st tokenunmeasured
- Z.ai: GLM 5z-ai/glm-5toolsjson205k contextIn / 1M$0.66up $0.6Out / 1M$2.112up $1.92Margin+10%1st tokenunmeasured
- Relace: Relace Apply 3relace/relace-apply-3256k contextIn / 1M$0.935up $0.85Out / 1M$1.375up $1.25Margin+10%1st tokenunmeasured
- MiniMax: MiniMax M1minimax/minimax-m1tools1M contextIn / 1M$0.605up $0.55Out / 1M$2.42up $2.20Margin+10%1st tokenunmeasured
- OpenAI: o3 Mini (batch)openai/o3-mini:batchtoolsjson200k contextIn / 1M$0.605up $0.55Out / 1M$2.42up $2.20Margin+10%1st tokenunmeasured
- OpenAI: o4 Mini (batch)openai/o4-mini:batchvisiontoolsjson200k contextIn / 1M$0.605up $0.55Out / 1M$2.42up $2.20Margin+10%1st tokenunmeasured
- DeepSeek: DeepSeek V4 Pro 0813deepseek/deepseek-v4-pro-0813toolsjson1M contextIn / 1M$0.726up $0.66Out / 1M$2.178up $1.98Margin+10%1st tokenunmeasured
- AionLabs: Aion-2.0aion-labs/aion-2.0toolsjson131k contextIn / 1M$0.88up $0.8Out / 1M$1.76up $1.60Margin+10%1st tokenunmeasured
- AionLabs: Aion-RP 1.0 (8B)aion-labs/aion-rp-llama-3.1-8b33k contextIn / 1M$0.88up $0.8Out / 1M$1.76up $1.60Margin+10%1st tokenunmeasured
- Anthropic: Claude Haiku 4.5 (batch)anthropic/claude-haiku-4.5:batchvisiontoolsjson200k contextIn / 1M$0.55up $0.5Out / 1M$2.75up $2.50Margin+10%1st tokenunmeasured
- ByteDance Seed: Seed 2.1 Turbobytedance-seed/seed-2-1-turbovisiontoolsjson262k contextIn / 1M$0.55up $0.5Out / 1M$2.75up $2.50Margin+10%1st tokenunmeasured
- Nous: Hermes 3 405B Instructnousresearch/hermes-3-llama-3.1-405bjson131k contextIn / 1M$1.10up $1.00Out / 1M$1.10up $1.00Margin+10%1st tokenunmeasured
- Perplexity: Sonarperplexity/sonarvision127k contextIn / 1M$1.10up $1.00Out / 1M$1.10up $1.00Margin+10%1st tokenunmeasured
“up” is the upstream's published rate, read from its catalog. A tag with a tick was measured working by a probe on this deployment; one without is the upstream's own word. First token is the median measured here over the last week, from real traffic and probes. Prices follow the upstream: when it changes a rate, this list and the bill change with it.
Calculator
What a request costs
The three cheapest worlds in the list as it is filtered now. Add up to four of your own with Compare. Computed with the same function the gateway bills with. Token counts are yours to set; the tasks are round numbers to start from.
| World | Antimatter | Upstream | Margin |
|---|---|---|---|
| Mistral: Mistral Nemomistralai/mistral-nemo | $0.0366 | $0.0333 | $0.003330 |
| inclusionAI: Ling 3.0 Flash VLinclusionai/ling-3.0-flash-vl | $0.0514 | $0.0468 | $0.004676 |
| inclusionAI: Ling 3.0 Flashinclusionai/ling-3.0-flash | $0.0520 | $0.0473 | $0.004725 |
A thousand requests of 1,200 tokens in and 350 out. A cached prompt token costs less where the upstream publishes a cache rate.
By need
The cheapest world for a need
The worlds that read images, hold a long context, call tools or return JSON, ranked by what a typical message costs at these prices.
Autopilot
Let the router choose
An autopilot id picks a world per request and is billed at that world's price from this list, with the same margin. The answer names the world in x-antimatter-model.
antimatter/autoBalanced: price, measured speed and measured quality, scored together.antimatter/auto:cheapThe lowest blended price in the catalog.antimatter/auto:fastThe lowest measured time to first token.antimatter/auto:codeTool calls measured working and a context sized for code.antimatter/auto:longA long context window, from the catalog.antimatter/auto:visionA world that takes images as input.antimatter/auto:toolsA world whose tool calls were measured working.antimatter/auto:jsonA world whose structured output was measured holding.