Pick 1-2 models, enter one prompt, and compare the real outputs side by side — no need to juggle multiple platforms. Free, with a daily limit.
Choose models to compare (up to 2)
GPT-5.6 Luna
Claude Haiku 4.5
GPT-5.4 Mini
These are value-tier models set aside for the free trial; sign up to call every model in the catalog.
Q: Is this free? Is there a usage limit?
A: Completely free, no signup needed. Each IP has a daily cap to prevent abuse as a free inference proxy — plenty for normal trying-it-out use. For heavier use, sign up and call the real API, which isn't subject to this limit.
Q: Why can I only pick value-tier models, not flagship ones?
A: The free trial only opens value-tier models to keep costs in check. Flagship models (and reasoning, multimodal, and other tiers) are available normally once you sign up and use the dashboard chat or API — no restriction there.
Q: Is the output truncated?
A: Yes — the trial caps output tokens so you get a sense of the model's style without a full response. Real usage (signed-in chat or API) has no such cap; output length follows the model's own limits and whatever max_tokens you set.
Claude / GPT / Gemini — all accessible via one account. Free credits on signup.