Perplexity Agent
CoverageRelease Notes
Editor's note (June 1, 2026): We updated this post to correct GPT-5.2 instead of GPT-5.2-Codex being the exception for Copilot Code Review.
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Model details
GPT-5.2
OpenAI introduced GPT-5.2 on December 11, 2025 as a new frontier model series explicitly aimed at professional knowledge work and long-running agents, positioning it as the most capable release yet in that vein. The company frames the model around practical output: creating spreadsheets, building presentations, writing code, interpreting images, working with long contexts, using tools, and sustaining complex multi-step projects. On the GDPval benchmark measuring well-specified knowledge work tasks across 44 occupations, OpenAI reports that GPT-5.2 Thinking wins or ties 70.9% of comparisons, versus 38.8% for GPT-5.1 Thinking, and the announcement claims state-of-the-art results across several other evaluations, including software engineering and competition mathematics suites.
The release narrative emphasizes deeper logical chains, richer context handling, and agentic execution that produces shippable artifacts such as design documents, runnable code, unit tests, and deployment scripts with fewer iterations, reflecting a shift toward reliable reasoning partners for ambiguous, high-stakes enterprise workflows. The same announcement describes GPT-5.2 as built on a new architecture delivering superior performance and efficiency, though specific parameter counts, training corpus details, and architectural specifics are not disclosed in the available materials. For practitioners, the practical fit is knowledge-work automation, coding assistance with auditable outputs, multi-step research, and orchestrated agent pipelines where tool use and long-context retention matter more than casual conversation.
Perplexity Agentopenai/gpt-5.2gpt
Quick Info
Powered by- Provider
- Perplexity Agent
- Model key
- openai/gpt-5.2
- Release date
- Dec 11, 2025
- Last updated
- Dec 11, 2025
- Knowledge cutoff
- 2025-08-31
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $1.75
- Output token cost
- $14.00
Limits
- Input tokens
- 272,000 tokens
- Output tokens
- 128,000 tokens
- Context window
- 400,000 tokens
Transparent token rates
Compare GPT-5.2 pricing
Rates are shown per one million tokens. Combined means one million input plus one million output tokens.
Latest news about GPT-5.2
Perplexity Agent
Coverage
OpenAI is testing a new model for Codex called
Perplexity Agent
CoverageBenchmark
Discover how GPT-5.2 improves knowledge work with major upgrades in long-context reasoning, tool calling, coding, vision, and end-to-end workflow execution., Your weekly briefing on AI. Analysis and insights from the Senior Editors at DataCamp. Click to read The Median: A Data & AI Newsletter, a Substack publication wi
Perplexity Agent
CoverageBenchmark
Explore this breakdown of OpenAI’s GPT-5.2 performance across coding, reasoning, math, long-horizon planning, multimodal understanding, and tool use benchmarks to learn what results actually mean for building more powerful AI agents.
Azure
CoverageBenchmark
BenchLM's model record for GPT-5.2, marked "Established" and released December 11, 2025, reports a 65/100 capability score (rank 37 of 232) with data dated September 15, 2026. The profile lists a 400K token context window, roughly 80 tokens/second throughput with a 91.08s first-token latency, and API pricing of $1.75 p Across capability axes, the page places GPT-5.2 at the 32nd percentile for Agentic, 46th for Coding, 85th for Knowledge, and 53rd for Multimodal work, with several categories marked Not eligible or Not ranked. Its strongest eligible category is Multimodal & Grounded at rank 23, while Agentic sits lowest at rank 104 of
Perplexity Agent
CoverageBenchmark
GPT-5.2 is the latest frontier-grade model in the GPT-5 series, offering stronger agentic and long context perfomance compared to GPT-5.1. $1.75 per million input tokens, $14 per million output tokens. 400,000 token context window, maximum output of 128,000 tokens. Higher uptime with 2 providers. Includes independent b