Atlas

Models

← All models

GPT-5.3 Instant (source unspecified)

OpenAI

This benchmark entry retains GPT-5.3 Instant observations whose source does not identify the evaluated version behind the mutable gpt-5.3-chat-latest route. The March 3, 2026 launch and March 16 update have separate artifact rows; undated aliases and later cards' latest-version comparisons remain here unless their checkpoint can be established independently.

Benchmark scores

BenchmarkScore
Arena Text — Business, Management & Financial Ops (No Style Control)1384
Arena Text — Chinese (No Style Control)1432
Arena Text — Creative Writing (No Style Control)1355
Arena Text — English (No Style Control)1390
Arena Text — Entertainment, Sports & Media (No Style Control)1357
Arena Text — Exclude Ties (No Style Control)1368
Arena Text — Exclude Ties (Style Controlled)1454
Arena Text — Expert (No Style Control)1396
Arena Text — French (No Style Control)1398
Arena Text — French (Style Controlled)1471
Arena Text — German (No Style Control)1384
Arena Text — Hard Prompts (No Style Control)1399
Arena Text — Hard Prompts English (No Style Control)1396
Arena Text — Instruction Following (No Style Control)1378
Arena Text — Japanese (No Style Control)1352
Arena Text — Korean (No Style Control)1345
Arena Text — Legal & Government (No Style Control)1394
Arena Text — Life, Physical & Social Science (No Style Control)1385
Arena Text — Longer Query (No Style Control)1395
Arena Text — Medicine & Healthcare (No Style Control)1392
Arena Text — Multi-turn (No Style Control)1412
Arena Text — Non-English (No Style Control)1382
Arena Text — Non-English (Style Controlled)1444
Arena Text — Occupational Mathematical (No Style Control)1385
Arena Text — Polish (No Style Control)1396
Arena Text — Polish (Style Controlled)1475
Arena Text — Russian (No Style Control)1399
Arena Text — Software & IT Services (No Style Control)1408
Arena Text — Spanish (No Style Control)1371
Arena Text — Writing, Literature & Language (No Style Control)1378
Loading Atlas data…