[waleedlatif1]

Summary

• Add GPT-6.1 Terra with verified pricing, long-context tiers, reasoning and verbosity controls, and token limits.
• Regenerate the Agent streaming documentation.

Type of Change

• New feature

Testing

• bun run lint
• bun run type-check
• bun run check:audits (61 audits)
• bun run apps/sim/scripts/check-block-registry.ts origin/staging
• bun run --cwd apps/sim test providers/models.test.ts providers/utils.test.ts providers/index.test.ts providers/openai/ (379 tests)
• bun run test
• Checked hosted membership and pricing at and above the long-context threshold, including cache reads and writes.
• Verified against OpenAI model documentation, OpenAI pricing, and OpenRouter.

Checklist

• Code follows project style guidelines
• Self-reviewed my changes
• Tests added/updated and passing (new tests pass the test-audit authoring gate)
• No new warnings introduced
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated
docs	Ready	Preview	Oct 27, 2026 5:56pm UTC

[waleedlatif1]

Verification — gpt-6.1-terra

Field	Value	Source URL	Status
id	gpt-6.1-terra	OpenAI	✓ verified
contextWindow	1,050,000 tokens	OpenAI, OpenRouter	✓ verified
input	$1/M	OpenAI pricing, OpenRouter	✓ two sources agree
cachedInput	$0.05/M	OpenAI pricing, OpenRouter	✓ two sources agree
output	$5/M	OpenAI pricing, OpenRouter	✓ two sources agree
Long-context tier	Above 272K input tokens: $2 input, $0.10 cached input, $7.50 output per M	OpenAI	✓ documented multipliers applied
reasoningEffort	none, low, medium, high, xhigh, max	OpenAI	✓ verified
verbosity	low, medium, high	API reference, deployment guidance	✓ verified
maxOutputTokens	128,000	OpenAI	✓ verified
temperature	Omitted: unsupported with reasoning enabled	GPT-6 guide	✓ correctly omitted
releaseDate	2026-10-26	OpenAI changelog, OpenRouter	✓ verified
hosted/billing	Hosted; included by the existing OpenAI provider expansion	getHostedModels() and shouldBillModelUsage()	— existing provider policy; runtime membership confirmed

Disagreements

• None for the registered fields.

Unverified fields

• None among the added registry fields. Deployment credentials were not inspected; Terra uses the existing OpenAI hosted-key configuration.

Validation — gpt-6.1-terra

Field	Repo	Live docs	Source URL	Status
id	gpt-6.1-terra	Same	OpenAI	✓
releaseDate	2026-10-26	Same	Changelog	✓
Lifecycle	Active	Listed as current	Model catalog	✓
input	$1/M	Same	OpenAI, OpenRouter	✓
output	$5/M	Same	OpenAI, OpenRouter	✓
cachedInput	$0.05/M	Same	OpenAI, OpenRouter	✓
pricing.updatedAt	2026-10-27	Fetched this session	OpenAI pricing	✓
Long-context tier	Above 272K: $2 / $0.10 / $7.50	2x input/cache and 1.5x output	OpenAI	✓
contextWindow	1,050,000	Same	OpenAI	✓
maxOutputTokens	128,000	Same	OpenAI	✓
temperature	Omitted	Rejected with reasoning enabled	GPT-6 guide	✓
reasoningEffort	none/low/medium/high/xhigh/max	Same; consumed by openai/core.ts	OpenAI	✓
verbosity	low/medium/high	Same; consumed by openai/core.ts	API reference, deployment guide	✓
thinking.levels/default	Omitted	OpenAI uses reasoning.effort	Reasoning guide	✓
Streamed reasoning	Inherited summaries	Summaries; raw reasoning withheld	Reasoning guide	✓
nativeStructuredOutputs	Omitted	Structured Outputs supported; provider handles schemas directly without this flag	OpenAI	✓
toolUsageControl	Inherited true	Function calling supported through Responses	OpenAI	✓
computerUse	Omitted	Vendor supports it; Sim's OpenAI computer-use routing does not expose Terra	providers/utils.ts	— correctly avoids advertising an unimplemented provider path
deepResearch	Omitted	General-purpose SKU	OpenAI	✓
memory: false	Omitted	Conversational model using Responses	GPT-6 guide	✓
recommended	Omitted	Existing flagship recommendation preserved	providers/models.ts	— policy
speedOptimized	Omitted	Luna remains the smallest/fastest tier	GPT-6 guide	✓
Hosting/billing	Included in hosted list	Existing OpenAI provider policy	getHostedModels() runtime probe	— confirmed
Hosted credentials	Existing provider configuration	No new credential mechanism required	providers/openai/index.ts	— deployment values not inspected

Findings

• No critical issues, warnings, suggestions, or unverified registry fields.
• Confirmed the provider uses Responses, consumes reasoning and verbosity, and already prices cache writes separately. Boundary probes confirmed the long-context tier applies only above 272K total input tokens.

Disagreements between sources

• None for the registered fields.

1 model checked · 0 critical · 0 warnings · 0 suggestions · 0 unverified registry fields.

[waleedlatif1]

@greptile

[waleedlatif1]

@cubic-dev-ai review this PR

[cubic-dev-ai]

@cubic-dev-ai review this PR

@waleedlatif1 I have started the AI code review. It will take a few minutes to complete.

[cubic-dev-ai]

No issues found across 2 files

Confidence score: 5/5

• Automated review surfaced no issues in the provided summaries.
• No files require special attention.

_{Turn on auto-fix | Re-trigger cubic}

[greptile-apps]

Confidence Score: 5/5

[Medium impact] The PR appears safe to merge; no actionable defects were found.

Summary

Adds gpt-6.1-terra to the OpenAI catalog.

• Sim's OpenAI model catalog now includes GPT-6.1 Terra.

No actionable issues were found.

_{Reviews (1) · Last reviewed commit: "feat(models): add GPT-6.1 Terra" · Reviewed by Greptile}