Stretch your Claude plan with Reason
Open Reason on the web or the Mac app
Sign in at reasonmachines.com, or download the Reason app for macOS from reasonmachines.com/download. Both open the same Sessions. Sending work to Reason needs a Pro or Max plan.
Connect the plans you already pay for
In Settings › Models, under Subscriptions, connect ChatGPT (Plus, Pro or Business), Claude (Pro or Max), Grok (SuperGrok or X Premium+) or GitHub Copilot, and add API keys if you have them. Work on those accounts is billed by their provider, not in Reason credits.
Pick the model per task
In the composer, choose the model for each Session: a smaller model for routine fixes, a larger one for hard changes. Each Session shows the model and account it ran on.
Keep going at a limit
When the account a turn runs on hits its limit, Reason moves the turn to your other connected accounts for that provider, then to Reason credits if your workspace can pay, and says so in the conversation.
Claude Code on a plan or the API
| Way to pay | Price | What to know |
|---|---|---|
| Claude Pro | $20 a month, or $17 billed annually | Five-hour and weekly limits, shared with claude.ai |
| Claude Max | $100 a month | 5x Pro usage; its API credits can't be used in Claude Code |
| Claude Ultra | $200 a month | 10x Pro weekly; no five-hour limit once verified |
| API, Opus 5.5 | $4 in / $20 out per million tokens | Anthropic: about $13 per developer per active day on average |
| API, Sonnet 5.5 | $2 in / $10 out per million tokens | Anthropic's docs say Sonnet handles most coding tasks |
| API, Haiku 5.5 | $0.10 in / $0.50 out per million tokens | Up to 100,000-token prompts; good for subagents |
From Anthropic's pricing pages and Claude Code docs, checked Nov 2, 2026. Anthropic puts typical API spend at $150 to $250 per developer a month, under $30 per active day for 90% of users. Cache reads cost a fraction of input, and Claude Code caches prompts automatically.
Cost tips from Claude Code's own docs
Pick the model for the job: Sonnet for most coding, Opus for hard architectural decisions, Haiku for subagents. Clear the conversation when you switch to unrelated work, and compact long ones with a focus. Lower the effort setting for simple tasks, since thinking tokens bill as output.
Keep CLAUDE.md short (Anthropic suggests under 200 lines), use plan mode before big changes to avoid rework, and run the usage view to see where tokens go. Agent teams use about seven times the tokens of a normal session.
One setting catches people out: if an ANTHROPIC_API_KEY environment variable is set, Claude Code uses it instead of your Pro or Max plan, and the work bills as API usage.
Reason: all your AI subscriptions in one coding agent
Use the subscriptions you already pay for (ChatGPT Plus/Pro/Business, Claude Pro/Max, SuperGrok or X Premium+, GitHub Copilot) or your own API keys in one coding agent. Hit your limit? Reason keeps going on your other connected accounts or Reason credits.
Reason isn't the lowest-priced agent on this page: Pro is $50 a month with 1,250 credits a week, and Max is $200 with 5,000; a credit is $0.04. Work on your own connected subscriptions or API keys is billed by that provider, not in Reason credits, so you don't pay twice for a plan you already have.
On cost: in Reason's published benchmark on GPT-6 Astra at xhigh effort, with the same tasks and graders, Reason 1.1 scored 65.1 against Codex's 62.5 on Coding Agent Index v1.5, at 22% lower cost per solved task. That's Reason's benchmark of its own harness, with the runner published to reproduce it.
Sources (checked Nov 2, 2026)
- Anthropic: Claude plans and pricing
- Anthropic: using Claude Code with your Pro or Max plan
- Anthropic: what is the Max plan?
- Anthropic: what is the Ultra plan?
- Claude Code docs: manage costs effectively
- Anthropic: Claude API pricing
- Reason: Reason 1.1 Coding Agent Index results (Oct 10, 2026)
- Reason docs: plans
Questions
What is the cheapest way to use Claude Code?
Claude Pro at $20 a month, or $17 a month billed annually, which includes Claude Code within Pro's limits. On the API, use Sonnet or Haiku for routine work instead of Opus.
Is Claude Code cheaper on a subscription or the API?
For regular use, usually the subscription: Anthropic puts average API use at about $13 per active day and $150 to $250 a month per developer, while Pro is $20 a month within its limits.
How do I reduce Claude Code costs?
Anthropic's docs suggest Sonnet for most tasks, clearing context between unrelated work, compacting long sessions, lower effort for simple tasks, a short CLAUDE.md, and plan mode before large changes.
Why is Claude Code billing my API account?
If an ANTHROPIC_API_KEY environment variable is set, Claude Code uses that key instead of your Claude plan, so the work bills as API usage.
Can I use Max's API credits for Claude Code?
No. Anthropic says the monthly API credits included with Max can't be used in Claude Code or the Claude apps.
What happens in Reason when my plan hits its limit?
Hit your limit? Reason keeps going on your other connected accounts or Reason credits. The turn moves to your other connected accounts for the same provider, then to Reason credits if your workspace can pay.