The full ClaudeAPI.cheap catalog. Claude, GPT, and Gemini through one Anthropic-compatible and one OpenAI-compatible endpoint.
The agent default — best at long, autonomous coding work.
Newest flagship. New tokenizer, improved reasoning, advanced coding.
Most powerful. Complex analysis, research, advanced coding.
Powerful flagship from late 2025. Strong reasoning at lower latency.
Best balance of speed and intelligence. Production ready.
Sonnet 4.5 (2025-09-29 snapshot). Reliable mid-tier reasoning.
Fastest & cheapest. High-volume, latency-sensitive apps.
Frontier reasoning, native tool use, broadest world knowledge.
OpenAI flagship. Top-tier reasoning, multimodal, broadest knowledge.
Standard GPT-5 tier. Strong general purpose at moderate cost.
Budget GPT-5. High-volume tasks, fast responses.
Code-specialized GPT-5. Optimized for software engineering.
Legacy GPT-5.2. Kept for pinned-version workloads.
1M+ context, multimodal-first, the long-doc specialist.
Gemini 3 Pro flagship. Long context, multimodal, top reasoning.
Gemini 2.5 Pro. 2M context, strong reasoning, very long-doc capable.
Fast & cheap Gemini. High-volume, latency-sensitive apps.
Cheapest Gemini tier. Bulk classification, simple tasks.
Gemini 3 Flash. Balanced speed + quality, newer than 2.5 series.