GPT-5.1 vs GPT-5.2
At a Glance
| Compare |  GPT-5.1OpenAI | GPT-5.2OpenAI |
|---|---|---|
| Pricing and Limits | ||
| Input priceFrom · USD / 1M tokens | $1.25Openai ↗ · Sep 3, 2026 | $1.75Openai ↗ · Sep 3, 2026 |
| Output priceFrom · USD / 1M tokens | $10.00Openai ↗ · Sep 3, 2026 | $14.00Openai ↗ · Sep 3, 2026 |
| Context windowMaximum documented tokens | 400K | 400K |
| Model facts checked | Sep 3, 2026View model evidence → | Sep 3, 2026View model evidence → |
Token prices are the lowest available sourced USD rates; input and output may use different providers. Cost ranking estimates output spend on LiveBench, not a full request bill. Ranking methodology →
Available Benchmarks
| Benchmark | GPT-5.1 | GPT-5.2 |
|---|---|---|
| ARC-AGI-1verified-v1-613c385b0c93 · verified_score · leader | 72.8385% of row best · percent · GPT-5.1 (Thinking, High) | 86.17100% of row best · percent · GPT-5.2 (XHigh) |
| ARC-AGI-2verified-v2-be40992860f9 · verified_score · leader | 17.6433% of row best · percent · GPT-5.1 (Thinking, High) | 52.91100% of row best · percent · GPT-5.2 (XHigh) |
| LMArena Document Arenadocument-2026-09-13-a64ca3b12b6e · arena_rating · statistical tie | 1,403.17100% of row best · rating · gpt-5.1; 95% CI [1393.82886201, 1412.50542979]; votes 8244; rank 42 | 1,401.12100% of row best · rating · gpt-5.2; 95% CI [1395.19508120, 1407.03614053]; votes 28212; rank 44 |
| LMArena Search Arenasearch-2026-10-13-a64ca3b12b6e · arena_rating · leader | 1,199.70100% of row best · rating · gpt-5.1-search; 95% CI [1194.51255947, 1204.89089087]; votes 63042; rank 20 | 1,171.4698% of row best · rating · gpt-5.2-search-non-reasoning; 95% CI [1165.94554809, 1176.98399625]; votes 79796; rank 27 |
| LMArena Text Arenatext-2026-10-15-a64ca3b12b6e · arena_rating · leader | 1,422.28100% of row best · rating · gpt-5.1; 95% CI [1418.72432020, 1425.83291138]; votes 44434; rank 110 | 1,412.1499% of row best · rating · gpt-5.2; 95% CI [1408.99454704, 1415.29029705]; votes 83855; rank 136 |
| LMArena Vision Arenavision-2026-10-09-a64ca3b12b6e · arena_rating · statistical tie | 1,234.81100% of row best · rating · gpt-5.1; 95% CI [1227.28917044, 1242.33289013]; votes 10818; rank 76 | 1,229.01100% of row best · rating · gpt-5.2; 95% CI [1223.02274809, 1235.00649862]; votes 20127; rank 79 |
| Overall ResultCounted from the protocol-matched rows above · 2 ties | 2 benchmark winsNo overall winner | 2 benchmark winsNo overall winner |
Third-party benchmark Only like-for-like primary-publisher results are shown; raw scores, relative scores, configuration, and token spend remain visible.
Side-by-Side Facts
| Field | GPT-5.1 | GPT-5.2 |
|---|---|---|
| Developer | OpenAI | OpenAI |
| Family | Gpt 5 1 | Gpt 5 2 |
| Model | GPT-5.1 | GPT-5.2 |
| Version | GPT-5.1 | GPT-5.2 |
| Lifecycle | active | active |
| Released | 2025-11-13 | 2025-12-11 |
| Knowledge cutoff | 2024-09-30 | 2025-08-31 |
| Input modalities | Text, Image | Text, Image |
| Output modalities | Text | Text |
| Context window | 400K | 400K |
| Total parameters | Unknown | Unknown |
| Active parameters | Unknown | Unknown |
| License | Unknown | Unknown |
| Open weights | No | No |
| API available | Yes | Yes |
| Self-hostable | No | No |
| Provider access | Openai (Standard), Openrouter (Standard) | Openai (Standard), Openrouter (Standard) |
| Capabilities | chat, generation, reasoning, structured_outputs, tools | chat, generation, reasoning, structured_outputs, tools |
GPT-5.1 Capabilities
GPT-5.2 Capabilities
Related Comparisons
All image comparisons →| A | Pair | B | Context |
|---|---|---|---|
Claude Sonnet 4.5Anthropic | vs | GPT-5.1OpenAI | cross-developer peersimage, text |
Seed 2.1 TurboByteDance Seed | vs | GPT-5.1OpenAI | cross-developer peersimage, text |
 Gemini 3.5 Flash-LiteGoogle DeepMind | vs | GPT-5.1OpenAI | cross-developer peerstext |
Gemini 3.6 FlashGoogle DeepMind | vs | GPT-5.1OpenAI | cross-developer peerstext |
Gemini 3.7 FlashGoogle DeepMind | vs | GPT-5.1OpenAI | cross-developer peerstext |
GPT-5.1OpenAI | vs | GPT-5.2 ProOpenAI | family variantsimage, text |
GPT-5.1OpenAI | vs | GPT-5.5OpenAI | family variantsimage, text |
GPT-5.1OpenAI | vs |  Qwen3 VL PlusQwen | cross-developer peersimage, text |
GPT-5.1OpenAI | vs | cross-developer peersimage, text | |
GPT-5.1OpenAI | vs | Grok 4.5xAI | cross-developer peersimage, text |
Primary Evidence
Sources and Freshness
Questions
GPT-5.1 vs GPT-5.2 FAQs
Is GPT-5.1 or GPT-5.2 better for coding?+
This comparison does not currently contain a protocol-matched coding benchmark for both GPT-5.1 and GPT-5.2, so Model Markets cannot name a coding leader from pricing, context size, or capability labels alone.
Which is cheaper, GPT-5.1 or GPT-5.2?+
GPT-5.1 is $1.25 and GPT-5.2 is $1.75 per million tokens, so GPT-5.1 is cheaper on this metric. GPT-5.1 is $10.00 and GPT-5.2 is $14.00 per million tokens, so GPT-5.1 is cheaper on this metric.
Which has a larger context window, GPT-5.1 or GPT-5.2?+
Both models have the same sourced context window: 400,000 tokens each.
Which performs better in benchmarks, GPT-5.1 or GPT-5.2?+
There is no overall benchmark winner: The verified common benchmarks do not produce a majority winner.
Can GPT-5.1 or GPT-5.2 be self-hosted?+
Both models have the same sourced self-hosting status: unsupported. GPT-5.1 is not marked open weight; GPT-5.2 is not marked open weight.
Can GPT-5.1 and GPT-5.2 understand images?+
GPT-5.1 is documented with image input; GPT-5.2 is documented with image input. This reflects supported input modalities, not vision quality.
Which can generate longer answers, GPT-5.1 or GPT-5.2?+
Both models have the same sourced maximum output: 128,000 tokens each.
Do GPT-5.1 and GPT-5.2 support reasoning and tool use?+
GPT-5.1: reasoning, tool calling, and image input. GPT-5.2: reasoning, tool calling, and image input. Feature support does not establish relative quality.
Which is available from more inference providers, GPT-5.1 or GPT-5.2?+
GPT-5.1 has 2 sourced provider routes; GPT-5.2 has 2, a tie.
Which offers better value, GPT-5.1 or GPT-5.2?+
There is no universal value winner. Compare the input and output prices above with the matched benchmark result for your workload: cheaper tokens can be offset by different quality, token usage, latency, or provider availability.