Compare AI models
All 35 models across 8 families, sorted by input price. Every one is reachable through the same OpenAI-compatible endpoint with a single key.
| Model | Family | Context | Max out | Input / 1M | Cached / 1M | Output / 1M |
|---|---|---|---|---|---|---|
| DeepSeek V4.1 Flashdeepseek-v4.1-flash | DeepSeek | 1M | 66K | $0.04 | $0.01 | $0.04 |
| GLM-5.3 Flashglm-5.3-flash | GLM | 1M | 66K | $0.056 | $0.0056 | $0.056 |
| GLM-5.3 Flash Uncensoredglm-5.3-flash-uncensored | GLM | 1M | 66K | $0.056 | $0.0056 | $0.056 |
| Claude Haiku 5.5claude-haiku-5-5 | Claude | 1M | 64K | $0.06 | $0.02 | $0.06 |
| DeepSeek V4 Flashdeepseek-v4-flash-0731 | DeepSeek | 1M | 66K | $0.06 | $0.0053 | $0.06 |
| MiMo V2.6 Flashmimo-v2.6-flash | MiMo | 600K | 66K | $0.06 | $0.004 | $0.06 |
| GPT-6 Lunagpt-6-luna | GPT | 1.1M | 128K | $0.1 | $0.02 | $0.1 |
| GLM-5.2glm-5.2 | GLM | 1M | 66K | $0.12 | $0.034 | $0.12 |
| GLM-5.3glm-5.3 | GLM | 1M | 66K | $0.12 | $0.034 | $0.12 |
| GLM-5.3 Uncensoredglm-5.3-uncensored | GLM | 1M | 66K | $0.12 | $0.034 | $0.12 |
| Grok 4.5grok-4.5 | Grok | 200K | 64K | $0.2 | $0.04 | $0.2 |
| Grok 4.6grok-4.6 | Grok | 500K | 66K | $0.2 | $0.04 | $0.2 |
| Grok 4.7grok-4.7 | Grok | 500K | 66K | $0.2 | $0.04 | $0.2 |
| MiMo V2.6 Promimo-v2.6-pro | MiMo | 600K | 66K | $0.2 | $0.006 | $0.2 |
| Gemini 4 Argongemini-4-argon | Gemini | 128K | 16K | $0.3 | $0.08 | $0.3 |
| GPT-6 Solgpt-6-sol | GPT | 1.1M | 128K | $0.3 | $0.08 | $0.3 |
| GPT-6.1 Solgpt-6.1-sol | GPT | 1.1M | 128K | $0.3 | $0.08 | $0.3 |
| Kimi K3kimi-k3 | Kimi | 1M | 66K | $0.3 | $0.06 | $0.3 |
| GLM-5.3 Flash Abliteratedglm-5.3-flash-abliterated | GLM | 1M | 66K | $0.44 | $0.12 | $0.44 |
| DeepSeek V4 Prodeepseek-v4-pro-0813 | DeepSeek | 1M | 66K | $0.5 | $0.1 | $0.5 |
| DeepSeek V4 Pro Uncensoreddeepseek-v4-pro-uncensored | DeepSeek | 1M | 66K | $0.5 | $0.1 | $0.5 |
| GPT-5.5gpt-5.5 | GPT | 1M | 128K | $0.9 | $0.08 | $0.9 |
| GPT-5.6 Lunagpt-5.6-luna | GPT | 1M | 128K | $0.9 | $0.08 | $0.9 |
| GPT-5.6 Solgpt-5.6-sol | GPT | 1M | 128K | $0.9 | $0.08 | $0.9 |
| GPT-5.6 Terragpt-5.6-terra | GPT | 1M | 128K | $0.9 | $0.08 | $0.9 |
| GPT-6 Astragpt-6-astra | GPT | 1.1M | 128K | $0.9 | $0.08 | $0.9 |
| Claude Fable 5claude-fable-5 | Claude | 1M | 64K | $0.9 | $0.08 | $0.9 |
| Claude Opus 4.6claude-opus-4-6 | Claude | 1M | 64K | $0.9 | $0.08 | $0.9 |
| Claude Opus 4.7claude-opus-4-7 | Claude | 1M | 64K | $0.9 | $0.08 | $0.9 |
| Claude Opus 4.8claude-opus-4-8 | Claude | 1M | 64K | $0.9 | $0.08 | $0.9 |
| Claude Opus 5claude-opus-5 | Claude | 1M | 64K | $0.9 | $0.08 | $0.9 |
| Claude Sonnet 5claude-sonnet-5 | Claude | 1M | 64K | $0.9 | $0.08 | $0.9 |
| Claude Sonnet 5.5claude-sonnet-5-5 | Claude | 1M | 64K | $1 | $0.12 | $1 |
| Claude Opus 5.5claude-opus-5-5 | Claude | 1M | 128K | $1.2 | $0.13 | $1.2 |
| GLM-5.3 Abliteratedglm-5.3-abliterated | GLM | 1M | 66K | $1.5 | $0.44 | $1.5 |
GLM-5.3 Uncensored vs GLM-5.3 Abliterated
The uncensored tier relaxes the provider's refusals; the abliterated tier removes the refusal direction from the weights. Same family, same endpoint — pick abliterated when you need the change to hold across every request, uncensored when you want the standard weights with softer behaviour.
DeepSeek V4.1 Flash vs GLM-5.3 Flash
Both are low-cost, high-throughput workhorses. DeepSeek V4.1 Flash is the cheapest input rate in the catalog; GLM-5.3 Flash is the flash tier of a strong bilingual generalist. Route bulk traffic to whichever your prompts favour and escalate on failure.
Claude Opus 5.5 vs GPT-6.1 Sol
The two reasoning ceilings here. Claude for long-horizon reasoning, structured writing and code review; GPT-6.1 Sol for tool use and agentic workflows. Both carry 1M-token context.