GPT-5
OpenAI's flagship general-purpose model. Superseded by GPT-5.5 in April 2026.
Specifications
- Context window
- 400,000 tokens
- Modality
- text, vision, audio
- License
- proprietary
- Family
- GPT
- Release date
- 2025-08-07
Links
Provider status
Timeline
Released
Initial public availability.
Pricing changes, lineage updates, and new benchmark results appear here as they happen. See the releases feed for the latest vendor activity.
API pricing
| Tier | Input / 1M tokens | Output / 1M tokens | Cache read | As of |
|---|---|---|---|---|
| standard | $5.00 | $30.00 | — | 2025-08-07 · source ↗ |
Looking for consumer subscriptions? See OpenAI's plans →
Benchmarks
Model Index →| Benchmark | Score | Setting | Measured | Source |
|---|---|---|---|---|
| MMLU-Pro general knowledge | 85.1% | 5-shot CoT | 2025-08-07 | source ↗ |
| GPQA Diamond reasoning | 87.3% | 0-shot CoT | 2025-08-07 | source ↗ |
| HumanEval coding | 95.2% | pass@1 | 2025-08-07 | source ↗ |
| Aider Polyglot coding | 85.4% | edit + test | 2025-09-01 | source ↗ |
| AIME 2025 math | 91.0% | CoT max-thinking | 2025-08-07 | source ↗ |
| LiveCodeBench coding | 74.5% | pass@1, post-cutoff | 2025-09-15 | source ↗ |
| MMMU multimodal | 74.1% | val | 2025-08-07 | source ↗ |
Want to see how GPT-5 ranks across these? Open the Model Index leaderboard →
Infrastructure context
All intelligence →Compute, silicon, and capex events that shape GPT-5's economics.
- Data center
Microsoft Fairwater Wisconsin: dedicated 1.5GW OpenAI training campus
Microsoft's Mount Pleasant, WI "Fairwater" campus — disclosed as a dedicated OpenAI training facility — reached 1.5GW configuration by Q1 2026 with NVIDIA GB200 NVL72 racks. The site uses closed-loop water cooling to reduce evaporative loss. Total Microsoft AI capex run-rate for FY26 is ~$110B.
Accelerators: GB200Power: 1.50 GWLocation: Mount Pleasant, WI - Data center
Stargate Abilene: OpenAI + Oracle + SoftBank $500B compute pact, first 1.2GW campus
Stargate is a 4-year, $500B compute commitment announced January 2025, jointly funded by OpenAI, Oracle, SoftBank, and MGX. The flagship 1.2GW Abilene, TX campus came online in phases through 2025–2026, anchored by Oracle Cloud. Additional sites in Wisconsin and New Mexico are under construction. Powers OpenAI's training fleet for GPT-5.5 and beyond.
Power: 1.20 GWCapex: $500BLocation: Abilene, TX
Compare GPT-5 with…
Agents built on GPT-5
Related news
How GPT-5 helped immunologist Derya Unutmaz solve a 3-year-old mystery
GPT-5 Pro helped immunologist Derya Unutmaz solve a three-year-old mystery about T cell behavior. The result matters because it could advance cancer and autoimmune research by providing new insights into immune system regulation.
Daybreak: Tools for securing every organization in the world
OpenAI introduced Daybreak tools, including Codex Security and GPT-5.5-Cyber, to help organizations find, validate, and patch vulnerabilities at scale. The new tools are meant to bring AI-assisted security workflows to every organization, with a focus on faster vulnerability discovery and remediation.
Who decides when AI is too dangerous?
The Trump administration imposed export controls on Anthropic’s new Fable 5 model and the underlying Mythos model, then Anthropic took both offline because it said it could not reliably block access for foreign nationals, including employees in the U.S. Fable 5 was still unavailable in Claude days later, and the dispute is now a test case for whether U.S. AI regulation will function as a real safety framework or as a political weapon against companies that don’t comply.
Improving health intelligence in ChatGPT
GPT-5.5 Instant improves ChatGPT’s health and wellness responses with stronger reasoning, better context handling, clearer communication, and physician-informed evaluations. The update matters because it is aimed at making health advice in ChatGPT more reliable and easier to understand, which is especially important for sensitive medical and wellness queries.
A near-autonomous AI chemist improves a challenging reaction in medicinal chemistry
OpenAI and Molecule.one reported that a near-autonomous AI chemist using GPT-5.4 improved a challenging medicinal-chemistry reaction. The result suggests large models can do more than propose molecules, potentially accelerating optimization of real drug-making steps.
How engineers at Nextdoor use Codex to build without limits
Nextdoor engineers are using Codex with GPT-5.5 to investigate hard-to-reproduce issues and build across platforms. The setup is meant to remove engineering bottlenecks so they can spend more time on product outcomes instead of debugging and platform-specific work.
Microsoft’s AI chief says superintelligence is near, but won’t take your job
Microsoft AI CEO Mustafa Suleyman said Microsoft’s October renewal with OpenAI freed it to pursue superintelligence independently, and that he has since assembled a Superintelligence team, built new training clusters, and announced seven new models at Build. He argued the OpenAI partnership remains one of the most successful in history, while framing Microsoft’s move as a shift toward frontier model development rather than a focus on taking people’s jobs.
How Wasmer used Codex to build a Node.js runtime for the edge
Wasmer used Codex with GPT-5.5 to build a Node.js runtime for the edge, accelerating development by 10x to 20x and shipping in weeks instead of months. The notable detail is that an AI-assisted workflow compressed a project that would normally take months into weeks, highlighting how Codex can speed up infrastructure-level development.