# The Token Bill — AI coding deals leaderboard

> Live pricing and SWE-bench scores for AI APIs used in coding harnesses. Prices refreshed daily.

- Last updated: 2026-08-25T21:33:56.443Z
- Providers tracked: 110
- Models listed: 61
- Total host offers: 1368
- Canonical HTML view: https://thetokenbill.com/

## Methodology

- **Input / Output $/MTok** — pay-per-token is direct; subscription offers are normalized to an effective rate assuming 5M input + 1M output tokens/day.
- **SWE-bench** — accuracy scores scraped from vals.ai's public SWE-bench leaderboard and fuzzy-matched to each model's canonical name.
- **Price/perf score** — `SWE-bench ÷ effective blended $/MTok`. Higher is better. Models with no SWE-bench score are ranked last.
- **Hosts** — most models are served by multiple hosts (e.g. Anthropic, Bedrock, Vertex; or DeepInfra, Together, Groq for open models). The main table shows the cheapest host per model and the host count; per-host breakdowns follow.

## Ranked models

| # | Model | Cheapest Host | Hosts | SWE-bench | Input $/MTok | Output $/MTok | Save | Type |
|---|-------|---------------|------:|----------:|-------------:|--------------:|-----:|------|
| 1 | DeepSeek V4 Flash 0731 | Relace | 32 | 88.8% | $0.04 | $0.08 | −92% | open |
| 2 | Gemini 3.1 Flash Lite Preview | Google AI Studio | 2 | 62.8% | $0.13 | $0.75 | −50% | closed |
| 3 | MiniMax M2.5 | SambaNova | 23 | 74.2% | $0.15 | $0.59 | −29% | closed |
| 4 | GPT-5.6 Luna | Azure | 3 | 93.0% | $0.20 | $1.20 | −9% | closed |
| 5 | MiniMax M3 | CoreWeave | 15 | 75.0% | $0.23 | $0.96 | −69% | closed |
| 6 | MiniMax M2.7 | Mara | 12 | 73.8% | $0.24 | $0.96 | −60% | closed |
| 7 | MiniMax M2.1 | AtlasCloud | 6 | 74.8% | $0.29 | $0.95 | −26% | closed |
| 8 | gpt-5.4-nano | OpenAI | 2 | 69.8% | $0.20 | $1.25 | — | closed |
| 9 | GLM 5.2 | Baidu | 35 | 82.8% | $0.49 | $1.54 | −78% | closed |
| 10 | DeepSeek V4 Pro 0813 | Alibaba | 15 | 96.4% | $0.58 | $1.74 | −60% | open |
| 11 | Qwen3.6 Plus | Alibaba | 1 | 73.4% | $0.33 | $1.95 | — | open |
| 12 | Qwen3.6 27B | Chutes | 13 | 70.0% | $0.29 | $2.00 | −47% | open |
| 13 | DeepSeek V4 Pro 0423 | StreamLake | 23 | 77.4% | $0.57 | $1.15 | −86% | open |
| 14 | Qwen3.8 27B | Alibaba | 10 | 86.0% | $0.35 | $2.55 | −19% | open |
| 15 | GLM 4.7 | DekaLLM | 16 | 69.4% | $0.38 | $1.74 | −74% | closed |
| 16 | Gemini 3.5 Flash Lite | Google AI Studio | 2 | 75.0% | $0.30 | $2.50 | — | closed |
| 17 | gpt-5-mini | OpenAI | 2 | 60.8% | $0.25 | $2.00 | — | closed |
| 18 | Kimi K2.5 | ModelRun | 20 | 70.0% | $0.40 | $1.90 | −35% | open |
| 19 | Kimi K2.6 | Decart | 28 | 76.2% | $0.55 | $2.31 | −52% | open |
| 20 | Nemotron 3 Ultra | DeepInfra | 6 | 69.0% | $0.50 | $2.20 | −29% | open |
| 21 | GLM 5 | GMICloud | 20 | 71.4% | $0.60 | $1.92 | −48% | closed |
| 22 | Devstral 2 2512 | Mistral | 1 | 62.8% | $0.44 | $2.20 | — | open |
| 23 | Gemini 3 Flash Preview | Google AI Studio | 2 | 75.0% | $0.50 | $3.00 | — | closed |
| 24 | Kimi K2 0711 | Novita | 2 | 60.2% | $0.57 | $2.30 | −6% | open |
| 25 | Kimi K2.7 Code | Inceptron | 21 | 78.2% | $0.67 | $3.40 | −23% | open |
| 26 | Gemini 3.7 Flash | Google AI Studio | 2 | 80.8% | $0.75 | $3.75 | — | closed |
| 27 | Gemini 3.6 Flash | Google AI Studio | 2 | 79.6% | $0.75 | $3.75 | — | closed |
| 28 | GLM 5.1 | Baidu | 27 | 76.4% | $0.90 | $2.82 | −42% | closed |
| 29 | GPT-5.4 Mini | Azure | 2 | 73.0% | $0.75 | $4.50 | — | closed |
| 30 | GLM 5.3 | Z Ai | 1 | 95.4% | $1.40 | $4.40 | — | closed |
| 31 | Muse Spark 1.2 | Meta | 1 | 86.6% | $1.25 | $4.25 | — | closed |
| 32 | Grok 4.3 | Xai | 1 | 71.4% | $1.25 | $2.50 | — | closed |
| 33 | Muse Spark 1.1 | Meta | 1 | 82.0% | $1.25 | $4.25 | — | closed |
| 34 | Gemini 3.1 Pro Preview | Google AI Studio | 2 | 78.8% | $1.00 | $6.00 | −50% | closed |
| 35 | Claude Haiku 4.5 | Anthropic | 4 | 66.6% | $1.00 | $5.00 | — | closed |
| 36 | Qwen3.6 Max Preview | Alibaba | 1 | 72.8% | $1.03 | $6.16 | — | open |
| 37 | Grok 4.6 | Xai | 1 | 95.6% | $2.00 | $6.00 | — | closed |
| 38 | Qwen3.7 Max | Alibaba | 1 | 68.8% | $1.48 | $4.42 | — | open |
| 39 | Grok 4.5 | Xai | 1 | 86.6% | $2.00 | $6.00 | — | closed |
| 40 | Qwen3.8 Max | Alibaba | 1 | 85.6% | $2.00 | $6.00 | — | open |
| 41 | Gemini 3.5 Flash | Google AI Studio | 2 | 78.8% | $1.50 | $9.00 | — | closed |
| 42 | GPT-5.6 Terra | Azure | 3 | 95.4% | $2.00 | $12.00 | −9% | closed |
| 43 | gpt-5.1 | OpenAI | 2 | 69.8% | $1.25 | $10.00 | — | closed |
| 44 | gpt-5 | OpenAI | 2 | 69.0% | $1.25 | $10.00 | — | closed |
| 45 | Claude Sonnet 5 | Anthropic | 5 | 79.6% | $2.00 | $10.00 | — | closed |
| 46 | Kimi K3 | Sail Research | 15 | 93.4% | $2.60 | $13.00 | −25% | open |
| 47 | GPT-5.3-Codex | OpenAI | 2 | 78.0% | $1.75 | $14.00 | — | closed |
| 48 | Gemini 2.5 Pro | Google AI Studio | 2 | 54.4% | $1.25 | $10.00 | — | closed |
| 49 | gpt-5.2 | OpenAI | 2 | 75.8% | $1.75 | $14.00 | — | closed |
| 50 | GPT-5.2-Codex | OpenAI | 2 | 72.4% | $1.75 | $14.00 | — | closed |
| 51 | gpt-5.4 (<272K context length) | OpenAI | 3 | 78.2% | $2.50 | $15.00 | −9% | closed |
| 52 | Claude Sonnet 4.6 | Claude Platform on AWS | 5 | 77.4% | $3.00 | $15.00 | — | closed |
| 53 | gpt-5.6-sol | OpenAI | 3 | 96.2% | $4.00 | $20.00 | −34% | closed |
| 54 | Claude Sonnet 4.5 | Google AI Studio | 5 | 70.0% | $3.00 | $15.00 | — | closed |
| 55 | Claude Opus 5 | Claude Platform on AWS | 5 | 97.0% | $5.00 | $25.00 | — | closed |
| 56 | Claude Opus 4.8 | Google AI Studio | 5 | 88.6% | $5.00 | $25.00 | — | closed |
| 57 | Claude Opus 4.7 | Amazon Bedrock | 5 | 82.0% | $5.00 | $25.00 | — | closed |
| 58 | Claude Opus 4.6 | Azure | 5 | 78.2% | $5.00 | $25.00 | — | closed |
| 59 | Claude Opus 4.5 | Claude Platform on AWS | 5 | 76.4% | $5.00 | $25.00 | — | closed |
| 60 | gpt-5.5 (<272K context length) | OpenAI | 3 | 82.6% | $5.00 | $30.00 | −9% | closed |
| 61 | Claude Fable 5 | Amazon Bedrock | 5 | 95.0% | $10.00 | $50.00 | — | closed |

## Host breakdown

### DeepSeek V4 Flash 0731

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Relace | $0.04 | $0.08 | fp4 | 98.7% |
| OpenInference | $0.04 | $0.10 | fp4 | 100.0% |
| Baidu | $0.06 | $0.12 | fp8 | 96.3% |
| GMICloud | $0.06 | $0.13 | fp4 | 99.0% |
| Decart | $0.06 | $0.13 | fp4 | 63.1% |
| Sail Research | $0.07 | $0.18 | fp4 | 98.6% |
| DeepInfra | $0.08 | $0.18 | fp8 | 99.7% |
| Ambient | $0.08 | $0.18 | fp4 | 99.6% |
| DigitalOcean | $0.08 | $0.25 | — | 99.7% |
| Morph | $0.08 | $0.28 | bf16 | 98.3% |
| BaseTen | $0.13 | $0.26 | fp8 | 99.3% |
| Inceptron | $0.13 | $0.28 | fp4 | 99.7% |
| CoreWeave | $0.13 | $0.28 | fp8 | 99.9% |
| Nebius | $0.14 | $0.28 | fp8 | 77.3% |
| AkashML | $0.14 | $0.28 | fp8 | 99.1% |
| Parasail | $0.14 | $0.28 | fp8 | 99.7% |
| Together | $0.14 | $0.28 | — | 99.0% |
| SiliconFlow | $0.14 | $0.28 | fp8 | 99.3% |
| Venice | $0.17 | $0.35 | — | 99.5% |
| Mancer 2 | $0.15 | $0.45 | fp8 | 99.4% |
| Phala | $0.20 | $0.40 | — | 99.3% |
| Ionstream | $0.21 | $0.42 | fp4 | 99.1% |
| Alibaba | $0.18 | $0.53 | — | 100.0% |
| Io Net | $0.22 | $0.49 | fp8 | 99.9% |
| StreamLake | $0.19 | $0.57 | fp8 | 99.3% |
| Wafer | $0.28 | $0.56 | — | 99.8% |
| Fireworks AI | $0.22 | $0.66 | — | 99.7% |
| Reka | $0.22 | $0.66 | fp4 | 100.0% |
| DeepSeek | $0.22 | $0.66 | — | 100.0% |
| Novita | $0.44 | $1.32 | fp8 | 99.9% |
| AtlasCloud | $0.44 | $1.32 | fp4 | 99.5% |
| Cloudflare | $0.44 | $1.32 | — | 99.9% |

### Gemini 3.1 Flash Lite Preview

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $0.13 | $0.75 | — | 0.0% |
| Google AI Studio | $0.25 | $1.50 | — | 99.2% |

### MiniMax M2.5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| SambaNova | $0.22 | $0.59 | — | 99.7% |
| Inceptron | $0.22 | $0.90 | fp8 | 100.0% |
| Mara | $0.24 | $0.96 | — | 99.3% |
| Venice | $0.27 | $0.95 | — | — |
| AkashML | $0.15 | $1.15 | fp8 | — |
| DeepInfra | $0.15 | $1.15 | fp8 | 100.0% |
| Chutes | $0.15 | $1.20 | fp8 | 100.0% |
| Baidu | $0.27 | $1.08 | fp8 | 100.0% |
| StreamLake | $0.27 | $1.08 | — | 100.0% |
| Io Net | $0.29 | $1.14 | fp8 | — |
| AtlasCloud | $0.29 | $1.20 | fp8 | 100.0% |
| NextBit | $0.30 | $1.20 | fp8 | 99.5% |
| Together | $0.30 | $1.20 | — | — |
| WandB | $0.30 | $1.20 | fp8 | 100.0% |
| Parasail | $0.30 | $1.20 | fp8 | 100.0% |
| CoreWeave | $0.30 | $1.20 | fp8 | — |
| Nebius | $0.30 | $1.20 | fp4 | 100.0% |
| Novita | $0.30 | $1.20 | fp8 | 100.0% |
| Minimax | $0.30 | $1.20 | fp8 | 100.0% |
| DigitalOcean | $0.30 | $1.20 | — | 100.0% |
| SiliconFlow | $0.30 | $1.20 | fp8 | 100.0% |
| Friendli | $0.30 | $1.20 | — | 100.0% |
| Phala | $0.20 | $1.38 | — | — |

### GPT-5.6 Luna

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Azure | $0.20 | $1.20 | — | 90.5% |
| OpenAI | $0.20 | $1.20 | — | — |
| Amazon Bedrock | $0.22 | $1.32 | — | 100.0% |

### MiniMax M3

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| CoreWeave | $0.23 | $0.96 | fp4 | 100.0% |
| Morph | $0.26 | $1.02 | fp4 | 87.6% |
| DeepInfra | $0.28 | $1.10 | fp8 | 99.1% |
| SiliconFlow | $0.30 | $1.20 | fp8 | 98.0% |
| AtlasCloud | $0.30 | $1.20 | fp8 | 100.0% |
| Parasail | $0.30 | $1.20 | fp8 | 98.1% |
| Novita | $0.30 | $1.20 | fp8 | 99.6% |
| StreamLake | $0.30 | $1.20 | fp8 | 98.8% |
| Together | $0.30 | $1.20 | — | 99.9% |
| Venice | $0.30 | $1.20 | fp8 | 100.0% |
| Minimax | $0.30 | $1.20 | fp8 | 99.9% |
| Fireworks AI | $0.30 | $1.20 | — | — |
| SambaNova | $0.60 | $2.40 | — | 99.7% |
| GMICloud | $0.60 | $2.40 | fp8 | 100.0% |
| ModelRun | $0.75 | $3.00 | fp4 | 99.8% |

### MiniMax M2.7

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Mara | $0.24 | $0.96 | — | 83.5% |
| DeepInfra | $0.25 | $1.00 | fp8 | 100.0% |
| Novita | $0.27 | $1.08 | fp8 | 99.8% |
| DekaLLM | $0.26 | $1.20 | fp4 | 100.0% |
| Morph | $0.28 | $1.20 | — | — |
| Together | $0.30 | $1.20 | fp4 | 99.9% |
| AtlasCloud | $0.30 | $1.20 | fp8 | 100.0% |
| Minimax | $0.30 | $1.20 | fp8 | 99.8% |
| Fireworks AI | $0.30 | $1.20 | — | — |
| GMICloud | $0.30 | $1.20 | fp8 | 99.5% |
| Groq | $0.60 | $1.80 | — | 97.8% |
| SambaNova | $0.60 | $2.40 | — | — |

### MiniMax M2.1

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| AtlasCloud | $0.29 | $0.95 | fp8 | — |
| Nebius | $0.30 | $1.20 | fp8 | 100.0% |
| Fireworks AI | $0.30 | $1.20 | — | — |
| Novita | $0.30 | $1.20 | fp8 | — |
| Minimax | $0.30 | $1.20 | fp8 | — |
| Venice | $0.35 | $1.50 | — | — |

### gpt-5.4-nano

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $0.20 | $1.25 | — | — |
| Azure | $0.20 | $1.25 | — | 100.0% |

### GLM 5.2

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Baidu | $0.49 | $1.54 | fp8 | 92.3% |
| Ambient | $0.60 | $2.00 | fp8 | 98.1% |
| StreamLake | $0.69 | $2.18 | fp8 | 99.9% |
| Novita | $0.70 | $2.20 | fp8 | 100.0% |
| DigitalOcean | $0.70 | $2.20 | — | 99.8% |
| Decart | $0.68 | $2.28 | fp4 | 91.1% |
| DeepInfra | $0.75 | $2.40 | fp4 | 99.2% |
| Inceptron | $0.75 | $2.40 | fp4 | 99.9% |
| CoreWeave | $0.76 | $2.42 | fp4 | 100.0% |
| AkashML | $0.77 | $2.42 | fp8 | 99.5% |
| Sail Research | $0.50 | $3.15 | fp8 | 100.0% |
| Morph | $0.84 | $3.14 | fp4 | 94.0% |
| Alibaba | $0.97 | $3.04 | fp8 | 100.0% |
| Phala | $1.26 | $3.00 | fp8 | 100.0% |
| GMICloud | $1.05 | $3.30 | fp8 | 99.6% |
| Reka | $1.10 | $3.75 | fp8 | 99.2% |
| SiliconFlow | $1.19 | $3.74 | fp8 | 100.0% |
| DekaLLM | $1.10 | $4.00 | — | 96.0% |
| Chutes | $1.25 | $3.95 | fp4 | — |
| AtlasCloud | $1.26 | $3.96 | fp8 | 99.6% |
| Wafer | $1.26 | $3.96 | fp4 | 99.6% |
| Io Net | $1.26 | $4.40 | fp8 | 100.0% |
| WandB | $1.39 | $4.40 | fp4 | 99.7% |
| Ionstream | $1.40 | $4.40 | fp4 | 97.4% |
| Mistral | $1.40 | $4.40 | — | 100.0% |
| Fireworks AI | $1.40 | $4.40 | — | — |
| Cloudflare | $1.40 | $4.40 | — | 99.1% |
| Z Ai | $1.40 | $4.40 | fp8 | 99.8% |
| Parasail | $1.40 | $4.40 | fp4 | 100.0% |
| Together | $1.40 | $4.40 | — | 99.9% |
| Crusoe | $1.40 | $4.40 | fp8 | 99.9% |
| BaseTen | $1.40 | $4.40 | fp8 | 99.8% |
| Venice | $1.40 | $4.40 | fp8 | 98.7% |
| Friendli | $1.40 | $4.40 | — | 99.9% |
| ModelRun | $2.20 | $6.80 | fp4 | 92.5% |

### DeepSeek V4 Pro 0813

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Alibaba | $0.58 | $1.74 | — | 99.9% |
| DeepSeek | $0.66 | $1.98 | — | 100.0% |
| DeepInfra | $1.30 | $2.60 | fp8 | 35.6% |
| GMICloud | $1.12 | $3.37 | fp8 | 99.9% |
| Sail Research | $1.32 | $3.96 | fp4 | 99.9% |
| StreamLake | $1.32 | $3.96 | — | 99.7% |
| Together | $1.32 | $3.96 | — | 95.4% |
| Novita | $1.32 | $3.96 | fp8 | 100.0% |
| Parasail | $1.32 | $3.96 | fp8 | 99.6% |
| SiliconFlow | $1.32 | $3.96 | fp8 | 100.0% |
| BaseTen | $1.32 | $3.96 | fp4 | 100.0% |
| DigitalOcean | $1.32 | $3.96 | — | 99.8% |
| Cloudflare | $1.32 | $3.96 | — | 99.7% |
| Fireworks AI | $1.32 | $3.96 | — | 100.0% |
| Phala | $1.45 | $4.36 | — | 100.0% |

### Qwen3.6 27B

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Chutes | $0.30 | $2.00 | fp8 | 95.9% |
| DekaLLM | $0.29 | $2.40 | — | 99.8% |
| Morph | $0.29 | $2.40 | fp4 | 100.0% |
| Phala | $0.32 | $2.70 | — | 94.8% |
| Alibaba | $0.45 | $2.70 | — | 100.0% |
| Io Net | $0.39 | $2.89 | fp8 | 100.0% |
| SiliconFlow | $0.30 | $3.20 | fp8 | 98.5% |
| Ambient | $0.32 | $3.20 | — | — |
| DeepInfra | $0.32 | $3.20 | fp8 | 100.0% |
| Venice | $0.33 | $3.25 | fp8 | 100.0% |
| Groq | $0.60 | $3.00 | — | — |
| WandB | $0.60 | $3.60 | fp8 | 100.0% |
| CoreWeave | $0.60 | $3.60 | fp8 | 100.0% |

### DeepSeek V4 Pro 0423

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| StreamLake | $0.57 | $1.15 | fp8 | 98.2% |
| Baidu | $0.57 | $1.15 | fp8 | 87.8% |
| GMICloud | $0.58 | $1.74 | fp8 | 100.0% |
| DigitalOcean | $0.87 | $1.74 | — | 99.7% |
| Ionstream | $1.13 | $2.26 | fp4 | 98.8% |
| Wafer | $1.20 | $2.40 | fp4 | 90.2% |
| Cloudflare | $1.15 | $2.55 | — | — |
| CoreWeave | $1.15 | $2.55 | fp8 | 100.0% |
| DeepInfra | $1.30 | $2.60 | fp8 | 98.3% |
| Alibaba | $1.42 | $2.83 | fp8 | 100.0% |
| SiliconFlow | $1.50 | $3.13 | fp8 | 97.6% |
| Novita | $1.60 | $3.20 | fp8 | 100.0% |
| Venice | $1.65 | $3.30 | — | 94.6% |
| AtlasCloud | $1.68 | $3.38 | fp4 | 100.0% |
| WandB | $1.74 | $3.48 | fp8 | 99.3% |
| Together | $1.74 | $3.48 | — | 99.3% |
| BaseTen | $1.74 | $3.48 | fp4 | 100.0% |
| Fireworks AI | $1.74 | $3.48 | — | — |
| Parasail | $1.74 | $3.48 | fp8 | 98.9% |
| NextBit | $1.75 | $3.50 | fp8 | 98.4% |
| DeepSeek | $1.32 | $3.96 | — | — |
| Azure | $1.91 | $3.83 | — | 97.7% |
| Io Net | $4.45 | $5.50 | — | — |

### Qwen3.8 27B

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Alibaba | $0.42 | $2.55 | — | 100.0% |
| Chutes | $0.35 | $2.75 | fp8 | 99.4% |
| Phala | $0.40 | $3.00 | — | 96.9% |
| CoreWeave | $0.40 | $3.00 | fp8 | 99.9% |
| AkashML | $0.40 | $3.00 | bf16 | 97.3% |
| Cloudflare | $0.45 | $3.20 | — | 94.0% |
| Reka | $0.45 | $3.20 | fp8 | 99.6% |
| Venice | $0.45 | $3.20 | fp8 | 97.9% |
| Parasail | $0.45 | $3.20 | fp8 | 100.0% |
| Io Net | $0.48 | $3.40 | fp8 | 100.0% |

### GLM 4.7

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| DekaLLM | $0.38 | $1.74 | fp4 | — |
| Chutes | $0.39 | $1.75 | bf16 | 100.0% |
| DeepInfra | $0.40 | $1.75 | fp4 | 100.0% |
| StreamLake | $0.48 | $1.76 | fp8 | 91.9% |
| AtlasCloud | $0.52 | $1.85 | fp8 | 94.5% |
| Nebius | $0.40 | $2.00 | fp8 | 93.9% |
| Novita | $0.54 | $1.98 | fp8 | 95.2% |
| Parasail | $0.45 | $2.10 | fp8 | 99.1% |
| SiliconFlow | $0.45 | $2.20 | fp8 | — |
| Fireworks AI | $0.60 | $2.20 | — | — |
| Z Ai | $0.60 | $2.20 | fp4 | 86.3% |
| Google AI Studio | $0.60 | $2.20 | — | 100.0% |
| Mancer 2 | $0.65 | $2.50 | fp4 | 100.0% |
| Venice | $0.55 | $2.65 | fp4 | 100.0% |
| Phala | $0.85 | $3.30 | — | — |
| Cerebras | $2.25 | $2.75 | fp16 | 100.0% |

### Gemini 3.5 Flash Lite

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $0.30 | $2.50 | — | 99.9% |
| Google AI Studio | $0.30 | $2.50 | — | — |

### gpt-5-mini

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $0.25 | $2.00 | — | — |
| Azure | $0.25 | $2.00 | — | 100.0% |

### Kimi K2.5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| ModelRun | $0.40 | $1.90 | fp4 | 99.9% |
| Ambient | $0.40 | $1.98 | — | 99.1% |
| Chutes | $0.44 | $2.00 | int4 | 90.4% |
| Io Net | $0.45 | $2.00 | int4 | 99.5% |
| Inceptron | $0.44 | $2.20 | int4 | 99.8% |
| DeepInfra | $0.45 | $2.25 | fp4 | 99.7% |
| SiliconFlow | $0.45 | $2.25 | int4 | 100.0% |
| AtlasCloud | $0.49 | $2.50 | int4 | 99.1% |
| DigitalOcean | $0.50 | $2.70 | — | 98.0% |
| StreamLake | $0.54 | $2.70 | fp8 | 99.9% |
| Together | $0.50 | $2.80 | — | 98.7% |
| Parasail | $0.60 | $2.80 | int4 | 99.8% |
| Novita | $0.57 | $2.85 | — | 100.0% |
| BaseTen | $0.60 | $3.00 | fp4 | — |
| Cloudflare | $0.60 | $3.00 | — | 99.8% |
| Fireworks AI | $0.60 | $3.00 | — | — |
| Amazon Bedrock | $0.60 | $3.00 | — | 100.0% |
| Moonshot AI | $0.60 | $3.00 | int4 | 100.0% |
| Phala | $0.60 | $3.00 | — | — |
| Venice | $0.53 | $3.33 | — | 99.9% |

### Kimi K2.6

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Decart | $0.55 | $2.31 | fp4 | 99.5% |
| Baidu | $0.56 | $2.36 | fp4 | 100.0% |
| StreamLake | $0.60 | $2.52 | fp8 | 99.8% |
| Ambient | $0.55 | $3.20 | int4 | 99.2% |
| Inceptron | $0.57 | $3.39 | int4 | 99.8% |
| Chutes | $0.58 | $3.40 | int4 | 97.5% |
| CoreWeave | $0.65 | $3.41 | fp4 | 98.6% |
| ModelRun | $0.70 | $3.40 | fp4 | 99.9% |
| Io Net | $0.66 | $3.50 | int4 | 100.0% |
| SiliconFlow | $0.77 | $3.40 | fp8 | 99.6% |
| Crusoe | $0.70 | $3.50 | bf16 | 100.0% |
| Novita | $0.80 | $3.40 | — | 99.5% |
| DeepInfra | $0.75 | $3.50 | fp4 | 99.2% |
| Venice | $0.75 | $3.50 | int4 | 100.0% |
| Parasail | $0.75 | $3.50 | int4 | 98.0% |
| GMICloud | $0.85 | $3.60 | fp8 | 97.3% |
| AkashML | $0.95 | $4.00 | int4 | 99.2% |
| Nebius | $0.95 | $4.00 | int4 | 99.1% |
| WandB | $0.95 | $4.00 | fp4 | 89.7% |
| DigitalOcean | $0.95 | $4.00 | — | 100.0% |
| Cloudflare | $0.95 | $4.00 | — | — |
| Moonshot AI | $0.95 | $4.00 | int4 | 99.9% |
| AtlasCloud | $0.95 | $4.00 | int4 | 98.4% |
| Fireworks AI | $0.95 | $4.00 | — | — |
| BaseTen | $0.95 | $4.00 | fp4 | — |
| Sail Research | $1.00 | $4.00 | int4 | 100.0% |
| Phala | $1.09 | $4.60 | — | 87.8% |
| Together | $1.20 | $4.50 | — | 100.0% |

### Nemotron 3 Ultra

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| DeepInfra | $0.50 | $2.20 | fp4 | 100.0% |
| BaseTen | $0.60 | $2.40 | fp4 | 100.0% |
| Venice | $0.63 | $3.13 | fp8 | 100.0% |
| Nebius | $1.00 | $3.00 | fp4 | 100.0% |
| Crusoe | $1.00 | $3.20 | fp4 | — |
| Together | $0.60 | $3.60 | — | 98.9% |

### GLM 5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| GMICloud | $0.60 | $1.92 | fp8 | 99.5% |
| StreamLake | $0.60 | $1.92 | fp8 | 99.9% |
| DeepInfra | $0.60 | $2.08 | fp4 | 100.0% |
| Baidu | $0.70 | $2.24 | fp8 | 100.0% |
| Ambient | $0.72 | $2.30 | fp8 | 99.3% |
| Chutes | $0.95 | $2.55 | fp8 | — |
| SiliconFlow | $0.95 | $2.55 | fp8 | 100.0% |
| BaseTen | $0.95 | $3.15 | fp4 | — |
| AtlasCloud | $0.95 | $3.15 | fp8 | — |
| Nebius | $1.00 | $3.20 | fp4 | 99.1% |
| Fireworks AI | $1.00 | $3.20 | — | — |
| Together | $1.00 | $3.20 | — | 100.0% |
| Friendli | $1.00 | $3.20 | — | 99.9% |
| Parasail | $1.00 | $3.20 | fp8 | 99.4% |
| Venice | $1.00 | $3.20 | fp8 | 100.0% |
| Z Ai | $1.00 | $3.20 | fp8 | 99.8% |
| Novita | $1.00 | $3.20 | fp8 | 100.0% |
| Amazon Bedrock | $1.00 | $3.20 | — | 100.0% |
| DigitalOcean | $1.00 | $3.20 | — | 98.3% |
| Phala | $1.20 | $3.50 | — | — |

### Gemini 3 Flash Preview

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $0.50 | $3.00 | — | 99.5% |
| Google AI Studio | $0.50 | $3.00 | — | — |

### Kimi K2 0711

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Novita | $0.57 | $2.30 | fp8 | 100.0% |
| Moonshot AI | $0.60 | $2.50 | fp8 | 68.6% |

### Kimi K2.7 Code

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Inceptron | $0.67 | $3.40 | int4 | 99.7% |
| DeepInfra | $0.68 | $3.40 | fp4 | 100.0% |
| Decart | $0.68 | $3.41 | fp4 | 98.8% |
| Ambient | $0.69 | $3.49 | — | 99.6% |
| CoreWeave | $0.71 | $3.50 | int4 | 100.0% |
| AkashML | $0.75 | $3.50 | int4 | — |
| Venice | $0.75 | $3.50 | int4 | 100.0% |
| Parasail | $0.76 | $3.50 | int4 | 100.0% |
| ModelRun | $0.85 | $3.75 | fp4 | 99.5% |
| SiliconFlow | $0.86 | $3.80 | fp8 | — |
| Novita | $0.91 | $3.84 | int4 | — |
| WandB | $0.94 | $4.00 | int4 | 100.0% |
| Io Net | $0.95 | $4.00 | int4 | — |
| Nebius | $0.95 | $4.00 | fp4 | — |
| Fireworks AI | $0.95 | $4.00 | — | — |
| AtlasCloud | $0.95 | $4.00 | int4 | 100.0% |
| GMICloud | $0.95 | $4.00 | fp8 | — |
| Cloudflare | $0.95 | $4.00 | — | — |
| Moonshot AI | $0.95 | $4.00 | int4 | — |
| Together | $0.95 | $4.00 | — | 100.0% |
| Alibaba | $0.95 | $4.00 | fp8 | — |

### Gemini 3.7 Flash

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $0.75 | $3.75 | — | 98.0% |
| Google AI Studio | $0.75 | $3.75 | — | — |

### Gemini 3.6 Flash

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $0.75 | $3.75 | — | 99.6% |
| Google AI Studio | $0.75 | $3.75 | — | — |

### GLM 5.1

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Baidu | $0.90 | $2.82 | fp8 | 100.0% |
| GMICloud | $0.91 | $2.86 | fp8 | 98.7% |
| StreamLake | $0.97 | $3.04 | fp8 | 99.9% |
| Chutes | $0.98 | $3.08 | fp8 | 95.8% |
| Wafer | $1.00 | $3.20 | fp4 | 100.0% |
| DeepInfra | $1.05 | $3.50 | fp4 | — |
| Ionstream | $1.11 | $3.52 | fp8 | — |
| SiliconFlow | $1.19 | $3.74 | fp8 | 100.0% |
| Inceptron | $1.40 | $3.74 | fp8 | 100.0% |
| AtlasCloud | $1.26 | $3.96 | fp8 | 99.4% |
| Phala | $1.21 | $4.20 | — | 94.8% |
| Io Net | $1.30 | $4.20 | fp8 | 99.6% |
| Alibaba | $1.33 | $4.18 | fp8 | 100.0% |
| Morph | $1.40 | $4.20 | — | — |
| BaseTen | $1.30 | $4.30 | fp4 | 100.0% |
| DigitalOcean | $1.30 | $4.30 | — | 100.0% |
| Crusoe | $1.20 | $4.40 | fp8 | 100.0% |
| Novita | $1.38 | $4.40 | fp8 | 100.0% |
| Ambient | $1.40 | $4.40 | fp8 | 96.8% |
| Together | $1.40 | $4.40 | — | 98.0% |
| Fireworks AI | $1.40 | $4.40 | — | — |
| CoreWeave | $1.40 | $4.40 | fp8 | 100.0% |
| Nebius | $1.40 | $4.40 | fp8 | 96.1% |
| Z Ai | $1.40 | $4.40 | fp8 | 98.8% |
| Parasail | $1.40 | $4.40 | fp8 | 98.4% |
| Friendli | $1.40 | $4.40 | — | 100.0% |
| Venice | $1.54 | $4.84 | fp8 | 94.5% |

### GPT-5.4 Mini

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Azure | $0.75 | $4.50 | — | 100.0% |
| OpenAI | $0.75 | $4.50 | — | — |

### Gemini 3.1 Pro Preview

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $1.00 | $6.00 | — | 98.4% |
| Google AI Studio | $2.00 | $12.00 | — | 99.9% |

### Claude Haiku 4.5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Anthropic | $1.00 | $5.00 | — | — |
| Google AI Studio | $1.00 | $5.00 | — | 99.9% |
| Azure | $1.00 | $5.00 | — | — |
| Amazon Bedrock | $1.00 | $5.00 | — | 99.9% |

### Gemini 3.5 Flash

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $1.50 | $9.00 | — | 99.2% |
| Google AI Studio | $1.50 | $9.00 | — | — |

### GPT-5.6 Terra

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Azure | $2.00 | $12.00 | — | 99.9% |
| OpenAI | $2.00 | $12.00 | — | — |
| Amazon Bedrock | $2.20 | $13.20 | — | — |

### gpt-5.1

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $1.25 | $10.00 | — | — |
| Azure | $1.25 | $10.00 | — | 100.0% |

### gpt-5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $1.25 | $10.00 | — | — |
| Azure | $1.25 | $10.00 | — | 100.0% |

### Claude Sonnet 5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Anthropic | $2.00 | $10.00 | — | — |
| Claude Platform on AWS | $2.00 | $10.00 | — | 100.0% |
| Azure | $2.00 | $10.00 | — | 99.5% |
| Google AI Studio | $2.00 | $10.00 | — | 100.0% |
| Amazon Bedrock | $2.00 | $10.00 | — | 100.0% |

### Kimi K3

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Sail Research | $2.60 | $13.00 | fp4 | 95.5% |
| Morph | $2.80 | $14.00 | fp4 | 96.8% |
| DeepInfra | $2.85 | $14.25 | bf16 | 99.2% |
| DigitalOcean | $2.85 | $14.25 | — | 99.8% |
| Nebius | $3.00 | $15.00 | fp4 | 47.0% |
| Parasail | $3.00 | $15.00 | fp4 | — |
| Wafer | $3.00 | $15.00 | — | 65.7% |
| BaseTen | $3.00 | $15.00 | fp8 | 93.3% |
| Together | $3.00 | $15.00 | — | 99.8% |
| Moonshot AI | $3.00 | $15.00 | mxfp4 | 97.8% |
| Phala | $3.00 | $15.00 | — | — |
| Modal | $3.00 | $15.00 | mxfp4 | 99.7% |
| Chutes | $3.00 | $15.00 | mxfp4 | 99.7% |
| Fireworks AI | $3.00 | $15.00 | — | — |
| Alibaba | $3.45 | $17.25 | — | 94.7% |

### GPT-5.3-Codex

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $1.75 | $14.00 | — | 99.6% |
| Azure | $1.75 | $14.00 | — | 100.0% |

### Gemini 2.5 Pro

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $1.25 | $10.00 | — | 98.4% |
| Google AI Studio | $1.25 | $10.00 | — | — |

### gpt-5.2

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $1.75 | $14.00 | — | — |
| Azure | $1.75 | $14.00 | — | 100.0% |

### GPT-5.2-Codex

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $1.75 | $14.00 | — | — |
| Azure | $1.75 | $14.00 | — | 100.0% |

### gpt-5.4 (<272K context length)

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $2.50 | $15.00 | — | — |
| Azure | $2.50 | $15.00 | — | 99.9% |
| Amazon Bedrock | $2.75 | $16.50 | — | — |

### Claude Sonnet 4.6

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Claude Platform on AWS | $3.00 | $15.00 | — | 100.0% |
| Anthropic | $3.00 | $15.00 | — | — |
| Amazon Bedrock | $3.00 | $15.00 | — | 100.0% |
| Azure | $3.00 | $15.00 | — | — |
| Google AI Studio | $3.00 | $15.00 | — | 100.0% |

### gpt-5.6-sol

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $4.00 | $20.00 | — | — |
| Azure | $5.00 | $30.00 | — | 98.8% |
| Amazon Bedrock | $5.50 | $33.00 | — | — |

### Claude Sonnet 4.5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $3.00 | $15.00 | — | 99.8% |
| Azure | $3.00 | $15.00 | — | 99.1% |
| Claude Platform on AWS | $3.00 | $15.00 | — | — |
| Amazon Bedrock | $3.00 | $15.00 | — | 99.9% |
| Anthropic | $3.00 | $15.00 | — | — |

### Claude Opus 5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Claude Platform on AWS | $5.00 | $25.00 | — | 99.9% |
| Azure | $5.00 | $25.00 | — | 100.0% |
| Anthropic | $5.00 | $25.00 | — | — |
| Google AI Studio | $5.00 | $25.00 | — | 99.9% |
| Amazon Bedrock | $5.00 | $25.00 | — | 100.0% |

### Claude Opus 4.8

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Google AI Studio | $5.00 | $25.00 | — | 100.0% |
| Azure | $5.00 | $25.00 | — | — |
| Amazon Bedrock | $5.00 | $25.00 | — | 100.0% |
| Anthropic | $5.00 | $25.00 | — | — |
| Claude Platform on AWS | $5.00 | $25.00 | — | 100.0% |

### Claude Opus 4.7

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Amazon Bedrock | $5.00 | $25.00 | — | 99.9% |
| Google AI Studio | $5.00 | $25.00 | — | 99.3% |
| Claude Platform on AWS | $5.00 | $25.00 | — | 100.0% |
| Anthropic | $5.00 | $25.00 | — | — |
| Azure | $5.00 | $25.00 | — | — |

### Claude Opus 4.6

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Azure | $5.00 | $25.00 | — | — |
| Google AI Studio | $5.00 | $25.00 | — | 100.0% |
| Amazon Bedrock | $5.00 | $25.00 | — | 99.7% |
| Anthropic | $5.00 | $25.00 | — | — |
| Claude Platform on AWS | $5.00 | $25.00 | — | 100.0% |

### Claude Opus 4.5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Claude Platform on AWS | $5.00 | $25.00 | — | — |
| Azure | $5.00 | $25.00 | — | — |
| Anthropic | $5.00 | $25.00 | — | — |
| Amazon Bedrock | $5.00 | $25.00 | — | 99.6% |
| Google AI Studio | $5.00 | $25.00 | — | — |

### gpt-5.5 (<272K context length)

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| OpenAI | $5.00 | $30.00 | — | — |
| Azure | $5.00 | $30.00 | — | 100.0% |
| Amazon Bedrock | $5.50 | $33.00 | — | — |

### Claude Fable 5

| Host | Input $/MTok | Output $/MTok | Quant | Uptime 30m |
|------|-------------:|--------------:|-------|-----------:|
| Amazon Bedrock | $10.00 | $50.00 | — | — |
| Azure | $10.00 | $50.00 | — | — |
| Claude Platform on AWS | $10.00 | $50.00 | — | 100.0% |
| Google AI Studio | $10.00 | $50.00 | — | 100.0% |
| Anthropic | $10.00 | $50.00 | — | — |


## Subscription deals

_None currently tracked._
