Claude Opus 4.6 & 4.5
Best long-form reasoning, longest context. The model to ask when the question is dense or the answer needs to span pages of analysis.
Live · 13 models · One flat price
Every frontier model.One conversation.
Run GPT-5.2 and Claude Opus on the same prompt — switch mid-chat, same flat $0.015 per message. No per-provider API keys, no separate billing dashboards. Just the answer that's actually best for the task.
$0.015 per message · USDC on Base · No subscription
Capabilities
What you can do here.
GPT-5.2 + Codex
Top general performance. GPT-5.2 Codex is the strongest pure coder. Mini and Nano variants for high-throughput cheap calls.
Kimi K2.5
Long context, structured output, cheaper per token than the frontier closed models. Strong multilingual.
Switch mid-conversation
Ask Claude. Don't like the answer? Re-run with GPT-5.2 or Kimi in the same chat. No setup change, no extra cost.
Compare side by side
'Run both GPT-5.2 and Claude Opus on this prompt and tell me where they differ' — model-battle is a first-class capability.
Privacy by default
Sessions sandboxed. No cross-user context, no training on your prompts. Conway Tech's permissionless inference enforces isolation.
Try one
Real questions, instantly answered.
Click any query — Conway runs it the moment you sign in.
vs the alternative
Why Conway, not the dashboard.
| Feature | Their dashboard | Conway chat |
|---|---|---|
| Per-provider keys | OpenAI key + Anthropic key + Moonshot key | None — Conway abstracts |
| Pricing | Per-model token billing | Flat $0.015/message |
| Switching cost | Code change | Pick from a menu |
| Side-by-side compare | Build it yourself | Built-in |
| Skill stack | None — raw inference | Nansen, GMGN, Cloud, Bankr |
| Privacy | Provider's policy | Sandboxed, no training |
FAQ
Common questions.
Which AI models does Conway support?
+
13 frontier models, all at the same per-message price: GPT-5.2, GPT-5 Mini, GPT-5 Nano, GPT-5.2 Codex, GPT-4o, GPT-4o Mini, Claude Opus 4.6, Claude Opus 4.5, Claude Sonnet 4.5, Claude Haiku 4.5, Kimi K2.5, Kimi K2.5 (Nvidia), and GLM-5 (Nvidia).
Why would I switch models?
+
Different models have genuinely different strengths. Claude Opus is the best at long-form reasoning and analysis. GPT-5.2 Codex is the strongest coder. Kimi handles long-context cheaper. With Conway you can run the same prompt through two models in the same conversation, side by side, for comparison.
Is the price the same across models?
+
Yes — $0.015 per message regardless of model. The infrastructure absorbs the variance; you don't have to think about it. Switch mid-conversation, no setting changes, no extra cost.
Are my conversations private?
+
Yes. Sessions are sandboxed per user. No cross-user context. Inference runs through Conway Tech's permissionless infrastructure with strict isolation. Your prompts are not used for model training.
Can I use my own API key?
+
Not currently — Conway abstracts model providers behind the platform balance for simplicity and a flat price. If you need direct API access for production workloads, the Conway Compute API is the way (talk to us).
Explore other modes
Or explore
Ready to use every frontier model at once?
Sign up in seconds — no credit card. Deposit USDC on Base when you're ready.