# Mercato

> AI inference market. One API key for Claude, GPT, Gemini, DeepSeek, Qwen and 130 models. Requests are bought from the cheapest trusted seller on the open market at cost (no margin), minus a holder discount funded by $MERCATO trading fees.

## Integrate
- OpenAI-compatible base URL: https://api.mercatocoin.com/v1
- Anthropic-compatible base URL: https://api.mercatocoin.com
- Auth: "Authorization: Bearer $MERCATO_API_KEY" (the Anthropic surface also accepts "x-api-key")
- Read the key from the MERCATO_API_KEY environment variable. Never hard-code it.
- Get a key at https://mercatocoin.com/app/keys

## Endpoints
- POST /v1/chat/completions  (OpenAI Chat Completions, stream: true supported; send stream_options.include_usage for usage + billing in the last chunk)
- POST /v1/messages          (Anthropic Messages, stream: true supported)
- POST /v1/messages/count_tokens
- GET  /v1/models            (list price, live market price and, with a key, your discount)

## Model ids
Format is provider/model. Any model works on either endpoint. List prices are USD per 1M input/output tokens.
- anthropic/claude-fable-5.1 (Claude Fable 5.1, 1,000,000 ctx, list $10/$50)
- anthropic/claude-opus-5.5 (Claude Opus 5.5, 1,000,000 ctx, list $4/$20)
- anthropic/claude-sonnet-5.5 (Claude Sonnet 5.5, 1,000,000 ctx, list $2/$10)
- anthropic/claude-haiku-5.5 (Claude Haiku 5.5, 1,000,000 ctx, list $0.1/$0.5)
- openai/gpt-6-astra (GPT-6 Astra, 1,050,000 ctx, list $10/$50)
- openai/gpt-6.1-sol (GPT-6.1 Sol, 1,050,000 ctx, list $2/$10)
- openai/gpt-6-luna (GPT-6 Luna, 1,050,000 ctx, list $0.1/$0.5)
- google/gemini-4-argon (Gemini 4 Argon, 1,048,576 ctx, list $2/$10)
- google/gemini-3.8-flash (Gemini 3.8 Flash, 1,048,576 ctx, list $0.825/$4.125)
- deepseek/deepseek-v4-pro-0813 (DeepSeek V4 Pro, 1,048,576 ctx, list $0.66/$1.98)
- deepseek/deepseek-v4.1-flash (DeepSeek V4.1 Flash, 1,048,576 ctx, list $0.024/$0.375)
- qwen/qwen3.8-max-0902 (Qwen3.8 Max, 1,000,000 ctx, list $2/$6)
- qwen/qwen3.8-flash (Qwen3.8 Flash, 1,000,000 ctx, list $0.15/$0.47)
- x-ai/grok-4.7 (Grok 4.7, 500,000 ctx, list $2/$6)
- moonshotai/kimi-k3 (Kimi K3, 1,048,576 ctx, list $0.239/$11.936)
- z-ai/glm-5.3 (GLM 5.3, 1,048,576 ctx, list $0.079/$1.4)
- mistralai/mistral-large-4-0 (Mistral Large 4, 1,048,576 ctx, list $1.36/$4.18)
- anthropic/claude-sonnet-5 (Claude Sonnet 5, 1,000,000 ctx, list $2/$10)
- anthropic/claude-opus-4.7 (Claude Opus 4.7, 1,000,000 ctx, list $5/$25)
- anthropic/claude-sonnet-4.6 (Claude Sonnet 4.6, 1,000,000 ctx, list $3/$15)
- anthropic/claude-haiku-4.5 (Claude Haiku 4.5, 200,000 ctx, list $1/$5)
- openai/gpt-5.6-sol (GPT-5.6 Sol, 1,050,000 ctx, list $4/$20)
- openai/gpt-5.6-luna (GPT-5.6 Luna, 1,050,000 ctx, list $0.2/$1.2)
- openai/gpt-5.5 (GPT-5.5, 1,050,000 ctx, list $5/$30)
- openai/gpt-5-mini (GPT-5 Mini, 400,000 ctx, list $0.25/$2)
- openai/gpt-4o-mini (GPT-4o mini, 128,000 ctx, list $0.15/$0.6)
- google/gemini-3.7-flash (Gemini 3.7 Flash, 1,048,576 ctx, list $0.75/$3.75)
- google/gemini-3.1-pro-preview (Gemini 3.1 Pro, 1,048,576 ctx, list $2/$12)
- google/gemini-3.1-flash-lite (Gemini 3.1 Flash Lite, 1,048,576 ctx, list $0.25/$1.5)
- deepseek/deepseek-v3.2 (DeepSeek V3.2, 163,840 ctx, list $0.28/$0.42)
- z-ai/glm-5.3-flash (GLM 5.3 Flash, 1,048,576 ctx, list $0.0052/$0.1985)
- z-ai/glm-5.2 (GLM 5.2, 1,048,576 ctx, list $0.0341/$7)
- moonshotai/kimi-k2.6 (Kimi K2.6, 262,144 ctx, list $0.65/$3.41)
- qwen/qwen3-coder (Qwen3 Coder, 262,144 ctx, list $0.3/$1)
- x-ai/grok-4.3 (Grok 4.3, 1,000,000 ctx, list $1.25/$2.5)
- minimax/minimax-m2.5 (MiniMax M2.5, 204,800 ctx, list $0.27/$0.95)
- mistralai/mistral-small-3.2-24b-instruct (Mistral Small 3.2, 256,000 ctx, list $0.09375/$0.25)
- deepseek/deepseek-v4-flash (DeepSeek V4 Flash 0423, 1,048,576 ctx, list $0.091/$0.182)
- deepseek/deepseek-v4-pro (DeepSeek V4 Pro 0423, 1,048,576 ctx, list $0.4427/$0.8853)
- qwen/qwen3-vl-235b-a22b-instruct (Qwen3 VL 235B A22B Instruct, 262,144 ctx, list $0.21/$1.9)
- moonshotai/kimi-k2.5 (Kimi K2.5, 262,144 ctx, list $0.5/$2.5)
- google/gemma-4-31b-it (Gemma 4 31B, 262,144 ctx, list $0.2/$0.4)
- z-ai/glm-4.6 (GLM 4.6, 204,800 ctx, list $0.55/$2.2)
- openai/gpt-5.4 (GPT-5.4, 1,050,000 ctx, list $2.5/$15)
- z-ai/glm-5.1 (GLM 5.1, 204,800 ctx, list $1.4/$4.4)
- qwen/qwen3-235b-a22b-2507 (Qwen3 235B A22B Instruct 2507, 262,144 ctx, list $0.09/$0.55)
- z-ai/glm-4.7 (GLM 4.7, 204,800 ctx, list $0.54/$1.98)
- minimax/minimax-m3 (MiniMax M3, 1,048,576 ctx, list $0.3/$1.2)
- qwen/qwen3-next-80b-a3b-instruct (Qwen3 Next 80B A3B Instruct, 262,144 ctx, list $0.09/$1.1)
- z-ai/glm-4.7-flash (GLM 4.7 Flash, 200,000 ctx, list $0.0605/$0.4)
- google/gemma-3-27b-it (Gemma 3 27B, 131,072 ctx, list $0.08/$0.35)
- z-ai/glm-5 (GLM 5, 204,800 ctx, list $1/$3.2)
- google/gemini-3-flash-preview (Gemini 3 Flash, 1,048,576 ctx, list $0.5/$3)
- openai/gpt-5.4-mini (GPT-5.4 Mini, 400,000 ctx, list $0.75/$4.5)
- meta/muse-wonder (Muse Wonder, 1,048,576 ctx, list $4/$20)
- moonshotai/kimi-k2.7-code (Kimi K2.7 Code, 262,144 ctx, list $0.6712/$3.35)
- google/gemini-2.5-flash (Gemini 2.5 Flash, 1,048,576 ctx, list $0.3/$2.5)
- openai/gpt-5.6-terra (GPT-5.6 Terra, 1,050,000 ctx, list $2/$12)
- anthropic/claude-opus-4.6 (Claude Opus 4.6, 1,000,000 ctx, list $5/$25)
- minimax/minimax-m2.1 (MiniMax M2.1, 204,800 ctx, list $0.3/$1.2)
- openai/gpt-6-sol (GPT-6 Sol, 1,050,000 ctx, list $2/$10)
- google/gemma-4-26b-a4b-it (Gemma 4 26B A4B, 262,144 ctx, list $0.0765/$0.255)
- qwen/qwen3-coder-next (Qwen3 Coder Next, 262,144 ctx, list $0.12/$0.8)
- minimax/minimax-m2 (MiniMax M2, 204,800 ctx, list $0.255/$1.02)
- anthropic/claude-opus-4.8 (Claude Opus 4.8, 1,000,000 ctx, list $5/$25)
- anthropic/claude-opus-5 (Claude Opus 5, 1,000,000 ctx, list $5/$25)
- anthropic/claude-sonnet-4.5 (Claude Sonnet 4.5, 1,000,000 ctx, list $3/$15)
- deepseek/deepseek-v4.1-flash-1017 (DeepSeek V4.1 Flash 1017, 1,048,576 ctx, list $0.15/$0.6)
- deepseek/deepseek-v4-flash-0731 (DeepSeek V4 Flash 0731, 1,048,576 ctx, list $0.44/$1.32)
- z-ai/glm-5-turbo (GLM 5 Turbo, 202,752 ctx, list $1.2/$4)
- anthropic/claude-fable-5 (Claude Fable 5, 1,000,000 ctx, list $10/$50)
- openai/gpt-5.3-codex (GPT-5.3-Codex, 400,000 ctx, list $1.75/$14)
- qwen/qwen3-32b (Qwen3 32B, 131,072 ctx, list $0.08/$0.28)
- x-ai/grok-4.6 (Grok 4.6, 500,000 ctx, list $2/$6)
- moonshotai/kimi-k2-thinking (Kimi K2 Thinking, 262,144 ctx, list $0.6/$2.5)
- openai/gpt-4o (GPT-4o, 128,000 ctx, list $2.5/$10)
- google/gemini-2.5-pro (Gemini 2.5 Pro, 1,048,576 ctx, list $1.25/$10)
- minimax/minimax-m2.7 (MiniMax M2.7, 204,800 ctx, list $0.3/$1.2)
- qwen/qwen3.6-27b (Qwen3.6 27B, 262,144 ctx, list $0.25/$2)
- openai/gpt-5.4-nano (GPT-5.4 Nano, 400,000 ctx, list $0.2/$1.25)
- x-ai/grok-4.5 (Grok 4.5, 500,000 ctx, list $2/$6)
- openai/gpt-5.2 (GPT-5.2, 400,000 ctx, list $1.75/$14)
- moonshotai/kimi-k2 (Kimi K2 0711, 131,072 ctx, list $0.57/$2.3)
- openai/gpt-5-nano (GPT-5 Nano, 400,000 ctx, list $0.05/$0.4)
- anthropic/claude-opus-4.5 (Claude Opus 4.5, 200,000 ctx, list $5/$25)
- meta-llama/llama-3.3-70b-instruct (Llama 3.3 70B Instruct, 131,072 ctx, list $0.1/$0.32)
- qwen/qwen3.8-2.4t-a95b (Qwen3.8 2.4T A95B, 1,048,576 ctx, list $2/$6)
- qwen/qwen3-coder-30b-a3b-instruct (Qwen3 Coder 30B A3B Instruct, 262,144 ctx, list $0.07/$0.28)
- z-ai/glm-5v-turbo (GLM 5V Turbo, 202,752 ctx, list $1.2/$4)
- z-ai/glm-4.5 (GLM 4.5, 131,072 ctx, list $0.6/$2.2)
- google/gemma-3-12b-it (Gemma 3 12B, 131,072 ctx, list $0.05/$0.15)
- google/gemma-3-4b-it (Gemma 3 4B, 131,072 ctx, list $0.04/$0.05)
- meta-llama/llama-3.2-3b-instruct (Llama 3.2 3B Instruct, 131,072 ctx, list $0.0509/$0.335)
- nousresearch/hermes-3-llama-3.1-405b (Hermes 3 405B Instruct, 131,072 ctx, list $1/$1)
- z-ai/glm-4.5-air (GLM 4.5 Air, 131,072 ctx, list $0.13/$0.85)
- bytedance-seed/seed-2-1-turbo (Seed 2.1 Turbo, 262,144 ctx, list $0.5/$2.5)
- qwen/qwen3-235b-a22b-thinking-2507 (Qwen3 235B A22B Thinking 2507, 128,000 ctx, list $0.45/$3.5)
- qwen/qwen3.5-397b-a17b (Qwen3.5 397B A17B, 262,144 ctx, list $0.5/$3.6)
- qwen/qwen3-30b-a3b (Qwen3 30B A3B, 40,960 ctx, list $0.12/$0.5)
- thinkingmachines/inkling (Inkling, 524,288 ctx, list $1/$4.05)
- google/gemini-3.5-flash (Gemini 3.5 Flash, 1,048,576 ctx, list $1.5/$9)
- google/gemini-3.5-flash-lite (Gemini 3.5 Flash Lite, 1,048,576 ctx, list $0.3/$2.5)
- google/gemini-3.6-flash (Gemini 3.6 Flash, 1,048,576 ctx, list $0.75/$3.75)
- openai/gpt-oss-120b (gpt-oss-120b, 131,072 ctx, list $0.15/$0.6)
- openai/gpt-oss-20b (gpt-oss-20b, 131,072 ctx, list $0.018/$0.09)
- openai/gpt-5.6-terra-pro (GPT-5.6 Terra Pro, 1,050,000 ctx, list $2/$12)
- openai/gpt-5.6-luna-pro (GPT-5.6 Luna Pro, 1,050,000 ctx, list $0.2/$1.2)
- deepseek/deepseek-chat-v3.1 (DeepSeek V3.1, 163,840 ctx, list $0.25/$0.95)
- mistralai/mistral-large-2512 (Mistral Large 3, 262,144 ctx, list $0.5/$1.5)
- mistralai/devstral-2512 (Devstral 2, 262,144 ctx, list $0.4/$2)
- mistralai/mistral-small-2603 (Mistral Small 4, 262,144 ctx, list $0.15/$0.6)
- mistralai/ministral-14b-2512 (Ministral 3 14B, 262,144 ctx, list $0.2/$0.2)
- mistralai/ministral-8b-2512 (Ministral 3 8B, 262,144 ctx, list $0.15/$0.15)
- mistralai/ministral-3b-2512 (Ministral 3 3B, 131,072 ctx, list $0.1/$0.1)
- x-ai/grok-4.20 (Grok 4.20, 2,000,000 ctx, list $1.25/$2.5)
- x-ai/grok-4.20-multi-agent (Grok 4.20 Multi-Agent, 2,000,000 ctx, list $1.25/$2.5)
- x-ai/grok-build-0.1 (Grok Build 0.1, 256,000 ctx, list $1/$2)
- qwen/qwen3.8-27b (Qwen3.8 27B, 1,000,000 ctx, list $0.2/$2.5)
- qwen/qwen3.7-plus (Qwen3.7 Plus, 1,000,000 ctx, list $0.32/$1.28)
- qwen/qwen3.7-max (Qwen3.7 Max, 1,000,000 ctx, list $1.475/$4.425)
- qwen/qwen3.5-35b-a3b (Qwen3.5 35B A3B, 262,144 ctx, list $0.14/$1)
- qwen/qwen3.6-35b-a3b (Qwen3.6 35B A3B, 262,144 ctx, list $0.1/$0.95)
- qwen/qwen3.5-9b (Qwen3.5 9B, 262,144 ctx, list $0.17/$0.25)
- qwen/qwen3-vl-30b-a3b-thinking (Qwen3 VL 30B A3B Thinking, 262,144 ctx, list $0.29/$1)
- nvidia/nemotron-3-nano-30b-a3b (Nemotron 3 Nano 30B, 262,144 ctx, list $0.06/$0.24)
- nvidia/nemotron-3-super-120b-a12b (Nemotron 3 Super 120B, 262,144 ctx, list $0.3/$2.4)
- nvidia/nemotron-3-ultra-550b-a55b (Nemotron 3 Ultra 550B, 262,144 ctx, list $0.6/$2.4)
- minimax/minimax-m3.1-flash (MiniMax M3.1 Flash, 1,048,576 ctx, list $0.15/$0.6)
- inception/mercury-2 (Mercury 2, 128,000 ctx, list $0.25/$0.75)
- meta/muse-spark-open (Muse Spark Open, 1,048,576 ctx, list $1.25/$4.25)

## Billing
Response headers: x-mercato-list-cost, x-mercato-market-cost, x-mercato-discount, x-mercato-charged, x-mercato-pool-factor, x-mercato-upstream. JSON responses also carry a "mercato" object with the same values.
discount = tier cap x pool factor. The pool resets daily at 00:00 UTC.

## Holder tiers
- Visitor: no $MERCATO, discount cap 0%
- Ascent: >= 0.05% of $MERCATO supply, discount cap 10%
- Orbit: >= 0.1% of $MERCATO supply, discount cap 30%
- Core: >= 0.5% of $MERCATO supply, discount cap 60%
- Singularity: >= 1% of $MERCATO supply, discount cap 80%

## Errors
400 invalid_request, 401 invalid_api_key, 402 insufficient_balance, 404 model_not_found, 429 rate_limited (honour retry-after), 502 upstream_error, 503 overloaded.
Retry 429/502/503 with exponential backoff.

## Docs
- https://mercatocoin.com/docs
- https://mercatocoin.com/docs/quickstart
- https://mercatocoin.com/docs/api