Key points
- The Agent API now supports google/gemini-4-carbon with introductory pricing through November 30, 2026.
- Support for anthropic/claude-fable-5-2 has been added to the Agent API models list.
- A new ultra effort level allows for broad research tasks using anthropic/claude-opus-5-5.
Perplexity updated its API changelog to document new model support and effort levels for its Agent API. The page now lists google/gemini-4-carbon and anthropic/claude-fable-5-2 as available options for developers.reportedPerplexity added two new AI models and a high-effort setting to its developer tools. These additions expand the range of foundation models accessible through the platform.
The claim
The changelog entry states that the Agent API now supports google/gemini-4-carbon.reported Introductory pricing applies through November 30, 2026, at $2.00 per million input tokens and $10.00 per million output tokens.reported Cached input tokens cost $0.10 per million.reported The entry also confirms support for anthropic/claude-fable-5-2.reported
The response
Perplexity introduced a new ultra effort level, designated as [email protected].reported This setting targets broad research and data-gathering tasks where coverage is the priority. Runs using this level can take a long time and support background mode execution. The ultra level runs on anthropic/claude-opus-5-5 and has no frozen configuration.reported
What the evidence means
The Decisions API received updates for URL fetching and web grounding. Developers can pass a page URL to fetch text before searching, with a cost of $0.001 per fetch.reported Web grounding checks questions against the web, running up to 100 Fast Searches.reported Each search costs $0.001.reported The Python and TypeScript SDKs do not yet have a mode parameter.
In practice
Users can omit max_output_tokens when using any model, including Anthropic models. Each generation can produce up to 128,000 output tokens, capped by the model’s limit.reported New 2026-11-11.0 versions of effort levels use this default.reported Earlier pinned versions retain their configured budgets. The Analytics API credit-usage responses currently omit the Credit Source breakdown.
Why it matters
- Developers can integrate Gemini 4 Carbon and Claude Fable 5.2 into agent workflows without waiting for separate announcements.
- Automatic output-token limits simplify configuration by removing the need to manually set max_output_tokens for many use cases.
- Decisions API web grounding allows applications to verify questions against current web data before generating answers.
Topics: models, API, Perplexity, Anthropic, Google
Updated
Then / now / next
The Agent API supported fewer foundation models and required explicit max_output_tokens settings.
The API supports Gemini 4 Carbon, Claude Fable 5.2, and automatic output-token limits up to 128,000 tokens.
Developers should monitor for SDK updates that include the mode parameter for web grounding.
Story timeline
- Then The Agent API supported fewer foundation models and required explicit max_output_tokens settings.
- 10 Nov 2026 Perplexity API Changelog page updated:
- 11 Nov 2026 Perplexity API Changelog page updated:
- 12 Nov 2026 Gondola doubles revenue as its team builds more with Codex
- 12 Nov 2026 AI chatbots are recommending fake retail websites, study finds
- 12 Nov 2026 Get quick answers from Gemini on any website.
- 12 Nov 2026 Samsung's Android XR smart glasses hands-on: stylish with good privacy protections like local face redaction and nice Gemini features like real-time translation (Victoria Song/The Verge)
- 12 Nov 2026 Jensen Huang built an AI kingdom. Now he’s gone, will Nvidia’s ramparts be breached?
- 12 Nov 2026 You can force Gemini to forget about you now – how to edit its Memory
- 12 Nov 2026 Google Search gets a major upgrade with Gemini 4 Hydrogen
- 12 Nov 2026 Gemini 4 Hydrogen: everyday coding and agentic workflows for developers
- 12 Nov 2026 Introducing Gemini 4 Hydrogen
- 12 Nov 2026 Google announces Gemini 4 Hydrogen with ‘significant leaps’ in speed, intelligence
- 12 Nov 2026 Gemini 4 Hydrogen
- 12 Nov 2026 Gemini Subscriptions page updated: Includes Gemini 4 Hydrogen
- Now The API supports Gemini 4 Carbon, Claude Fable 5.2, and automatic output-token limits up to 128,000 tokens.
- Next Developers should monitor for SDK updates that include the mode parameter for web grounding.
The strongest doubt
Counterpoint from the sources
The Analytics API credit-usage responses currently omit the Credit Source breakdown, which may affect billing integration logic.
Ask the article
Local keyword search of this article; no generated answers.Did this article give you the evidence you needed?
Your choice stays on this device. No public tally.Terms in this story
Agent API
Perplexity's interface for building autonomous agents that can perform tasks.
Effort level
A setting that determines the computational intensity and coverage of a request.
Web grounding
A process that checks questions against the web to add evidence to the state.
Questions readers ask
What is the pricing for Gemini 4 Carbon in the Agent API?
Introductory pricing through November 30, 2026, is $2.00 per million input tokens and $10.00 per million output tokens.
Can I use the ultra effort level with background mode?
Yes, the ultra effort level supports background mode because runs can take a long time.
Do the Python and TypeScript SDKs support the web grounding mode?
No, the Perplexity Python library and TypeScript SDK do not have a mode parameter yet; send it in the request body.
Connected stories
Reading receipt
- The Agent API now supports google/gemini-4-carbon with introductory pricing through November 30, 2026.
- Support for anthropic/claude-fable-5-2 has been added to the Agent API models list.
- A new ultra effort level allows for broad research tasks using anthropic/claude-opus-5-5.