Claude Mythos Preview vs Claude Opus 4.8
Wins 3 of 4 areas
Coding · Reasoning · Images and charts
Wins 0 of 4 areas
—
Claude Mythos Preview is the stronger all-rounder.
Scores updated · 22 tests both models report · How we compare
Where each one wins
Tests won in each of the four areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software30Claude Mythos Preview3 of 3 tests
- ReasoningHard problems that need careful thinking20Claude Mythos Preview2 of 2 tests
- Images and chartsUnderstanding pictures, charts and video10Claude Mythos Preview1 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each
Images and charts rests on a single test.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Mythos Preview pulls ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+16points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+8.6points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+5.3points ahead
Where Claude Opus 4.8 pulls ahead
- Completes tasks by operating a computer desktopOSWorld-Verified+3.8points ahead
Every test, side by side
All 22 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Mythos Preview
- SWE-bench ProClaude Mythos Preview by 8.677.869.2+8.6
- SWE-bench VerifiedClaude Mythos Preview by 5.393.988.6+5.3
- SWE-bench MultilingualClaude Mythos Preview by 2.987.384.4+2.9
AgentsEven
- OSWorld-VerifiedClaude Opus 4.8 by 3.879.683.4+3.8
- BrowseCompClaude Mythos Preview by 2.686.984.3+2.6
ReasoningClaude Mythos Preview
- Humanity's Last ExamClaude Mythos Preview by 1664.748.7+16
- GPQA DiamondClaude Mythos Preview by 2.694.692+2.6
Images and chartsClaude Mythos Preview
- CharXiv (RQ)Claude Mythos Preview by 3.393.289.9+3.3
Other results14 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- SWE-bench MultimodalClaude Mythos Preview by 20.65938.4+20.6
- GraphWalks Parents 1M subsetClaude Mythos Preview by 12.295.583.3+12.2
- RiemannBenchClaude Mythos Preview by 94334+9
- HealthBench ProfessionalClaude Mythos Preview by 8.964.755.8+8.9
- Terminal-Bench 2.0Claude Mythos Preview by 7.48274.6+7.4
- HLE (with tools)Claude Mythos Preview by 6.864.757.9+6.8
- GraphWalks BFS 1M subsetClaude Mythos Preview by 6.274.368.1+6.2
- CyberGymClaude Mythos Preview by 4.383.178.8+4.3
- Legal Agent Benchmark, Full Public SetClaude Mythos Preview by 3.813.49.6+3.8
- BrowseComp (single-agent)Claude Mythos Preview by 3.687.984.3+3.6
- CharXiv Reasoning (With tools)Claude Mythos Preview by 2.692.589.9+2.6
- HealthBenchClaude Mythos Preview by 1.861.159.3+1.8
- GraphWalks — Parents task, 256K-token context subsettie99.999.3tie
- GraphWalks BFS 256Ktie85.785.9tie
Questions people ask
Which is better, Claude Mythos Preview or Claude Opus 4.8?
Claude Mythos Preview wins three of the four areas where both have results: coding, reasoning and images and charts. Claude Opus 4.8 wins none. They are level on agents.
Which is better for coding?
Claude Mythos Preview. It wins 3 of the 3 coding tests both models report; Claude Opus 4.8 wins none.
How do you compare the two?
We use the 22 benchmark tests both models have published scores on. The verdict counts the 8 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 14 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.