GPT-5.5 Instant (June 2026)
Jun 2026
GPT-5.5 Instant (June 2026)Compare inon
Benchmarks of GPT-5.5 Instant (June 2026)17
BenchmarkByScoreRank
Overall1
Knowledge3
Science3
Coding2
Agents4
Health & law1
Documents2
Legacy · saturated or no longer reported1
No benchmark matches
$0.57per task?
API price not publicprice
158 tok/sfirst token in 0.9 s?
400Kcontext
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperClaude Haiku 5.5maxwins 94%-63%$0.21Remove sampling parameters+5
OpenQwen3.8 Flashthinkingwins 94%-35%$0.37Context: 400K → 262K+6
CheaperDeepSeek V4.1 Flash 1017maxOpenwins 94%-54%$0.26Pass back reasoning_content · Chat Completions+4
CheaperMiMo v2.6 ProthinkingOpenwins 92%-77%$0.13Pass back reasoning_content+3CheaperGPT-6.1 Solxhighwins 92%-31%$0.39Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
Claude Haiku 5.5maxwins 94%-63%$0.21Remove sampling parameters+5Qwen3.8 FlashthinkingOpenwins 94%-35%$0.37Context: 400K → 262K+6DeepSeek V4.1 Flash 1017maxOpenwins 94%-54%$0.26Pass back reasoning_content · Chat Completions+4MiMo v2.6 ProthinkingOpenwins 92%-77%$0.13Pass back reasoning_content+3GPT-6.1 Solxhighwins 92%-31%$0.39Remove unsupported parameters+5ModelScoreCostPer taskWhat changes
Claude Haiku 5.5lowwins 54%-96%$0.024Remove sampling parameters+5GPT-6 Lunahighwins 64%-95%$0.029Function calling only with none · Chat Completions+3GPT-5.6 Lunahighwins 62%-92%$0.044MiMo v2.6 FlashthinkingOpenwins 91%-89%$0.062Pass back reasoning_content+3GPT-6.1 Terramediumwins 65%-83%$0.095ModelScoreCostPer taskWhat changes
Qwen3.8 FlashthinkingOpenwins 94%-35%$0.37Context: 400K → 262K+6DeepSeek V4.1 Flash 1017maxOpenwins 94%-54%$0.26Pass back reasoning_content · Chat Completions+4MiMo v2.6 ProthinkingOpenwins 92%-77%$0.13Pass back reasoning_content+3
GLM 5.3maxOpenwins 92%+253%$2.01No image input+4
Kimi K3maxOpenwins 91%+252%$2.00Sampling parameters are fixed+3ModelScoreCostPer taskWhat changes
GPT-6 Astramediumwins 90%+171%$1.54Monitoring can return HTTP 403+2GPT-6.1 Solmediumwins 83%-62%$0.21Remove unsupported parameters+5GPT-5.6 Solmediumwins 71%-11%$0.50GPT-6.1 Terramediumwins 65%-83%$0.095GPT-6 Solmediumwins 59%-57%$0.25GPT-6 Lunamediumsame-97%$0.017Function calling only with none · Chat Completions+3GPT-5.6 Terramediumwins 43%-68%$0.18Prompts with >272K input tokensGPT-5.5mediumwins 38%+35%$0.77Prompts with >272K input tokensAll OpenAI models →Benchmarks
all categoriesChat & Knowledge
Humanity's Last Exam19.9#127
AA Omniscience Index3.7#58
CritPt (physics)0#182
AA Omniscience · Accuracy44.4#36
GDP.pdf · AA all-pass20.2#26
MLCR-AA31.7#6
GPQA Diamond82.3#103
Coding
Terminal-Bench 4.012.6#36
SciCode52.5#41
Terminal-Bench 2.134.8#93
Agents & Tool Use
AutomationBench-AA36.4#48
GDPval-AA (work tasks) 2.1508.3#101
τ³-Banking · AA11.8#85
AA-Briefcase 1.1 · Elo1,086.3#52
FAQ
How much does GPT-5.5 Instant (June 2026) cost?›
There is no per-unit API price: API price not public.
Is GPT-5.5 Instant (June 2026) free to use?›
There is no free API.
How good is GPT-5.5 Instant (June 2026), and what is it best at?›
#127 of 336 in Chat & Knowledge, #36 of 134 in Coding, #48 of 135 in Agents & Tool Use, #26 of 135 in Vision & Documents. Strongest at: Knowledge.
Is GPT-5.5 Instant (June 2026) better than MiniMax M3?›
On AA-Briefcase 1.1 · Elo, GPT-5.5 Instant (June 2026) scores 1,086.3 vs 1,089.5 — about the same. It is 12% more expensive per task.
What is the context window of GPT-5.5 Instant (June 2026)?›
400K tokens.
When was GPT-5.5 Instant (June 2026) released? Is there a newer version?›
Released 2026-06-25. It is the current version in its line.
Is GPT-5.5 Instant (June 2026) open source?›
No, GPT-5.5 Instant (June 2026) is a closed model available through an API. The closest open-weights option here is Qwen3.8 Flash.
Can GPT-5.5 Instant (June 2026) read images?›
It accepts text and images as input.
What is GPT-5.5 Instant (June 2026)?›
GPT-5.5 Instant (June 2026) is a closed language model from OpenAI, released Jun 2026, the successor to GPT-5.5 Instant (May 2026).
What are the best alternatives to GPT-5.5 Instant (June 2026)?›
Claude Haiku 5.5 (cheaper), Qwen3.8 Flash (open weights).