Claude Mythos 5 vs Claude Sonnet 5
Wins 4 of 5 areas
Coding · Agents · Reasoning · Math
Wins 0 of 5 areas
—
Claude Mythos 5 is the stronger all-rounder.Claude Sonnet 5 is cheaper.
Scores updated · 28 tests both models report · How we compare
Where each one wins
Tests won in each of the five 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 software40Claude Mythos 54 of 4 tests
- ReasoningHard problems that need careful thinking20Claude Mythos 52 of 2 tests
- AgentsCarrying out multi-step tasks on its own10Claude Mythos 51 of 1 test
- MathCompetition and research-level math10Claude Mythos 51 of 1 test
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 1 tie
Agents, images and charts and math rest on a single test each.
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.
Claude Sonnet 5 costs 80% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Mythos 5 pulls ahead
- Proof problems from the US Math OlympiadUSAMO 2026+20.3points ahead
- Long, multi-file coding tasks in real codebasesSWE-bench Pro+16.8points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+16.5points ahead
Where Claude Sonnet 5 pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 28 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Mythos 5
- SWE-bench ProClaude Mythos 5 by 16.88063.2+16.8
- SWE-bench VerifiedClaude Mythos 5 by 10.395.585.2+10.3
- SWE-bench MultilingualClaude Mythos 5 by 8.386.678.3+8.3
- Terminal-Bench 2.1Claude Mythos 5 by 7.58880.5+7.5
ReasoningClaude Mythos 5
- Humanity's Last ExamClaude Mythos 5 by 16.557.841.3+16.5
- CritPtClaude Mythos 5 by 11.728.616.9+11.7
Other results19 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.
- OSWorld 2.0 (partial)Claude Mythos 5 by 30.372.942.6+30.3
- Terminal-Bench 4.0Claude Mythos 5 by 29.64212.4+29.6
- SWE-bench MultimodalClaude Mythos 5 by 2654.128.1+26
- Toolathlon Pass@∞Claude Mythos 5 by 14.955.640.7+14.9
- GDPval-AA v2Claude Mythos 5 by 105 rating points17231618+105 rating
- ProteinGym HardClaude Mythos 5 by 9.245.836.6+9.2
- Program BenchClaude Mythos 5 by 986.377.3+9
- Legal Agent Benchmark, Full Public SetClaude Mythos 5 by 816.98.9+8
- HealthBench Professional (raw)Claude Mythos 5 by 7.970.362.4+7.9
- OfficeQA ProClaude Mythos 5 by 7.767.159.4+7.7
- ToolathlonClaude Mythos 5 by 6.861.154.3+6.8
- BrowseComp (multi-agent)Claude Mythos 5 by 6.793.386.6+6.7
- HLE (with tools)Claude Mythos 5 by 6.463.857.4+6.4
- GDP (Surge AI)Claude Mythos 5 by 5.787.381.6+5.7
- HealthBench ProfessionalClaude Mythos 5 by 5.563.357.8+5.5
- CharXiv Reasoning (With tools)Claude Mythos 5 by 5.293.588.3+5.2
- Toolathlon VerifiedClaude Mythos 5 by 4.679.374.7+4.6
- OfficeQAClaude Mythos 5 by 3.97975.1+3.9
- HealthBench (raw)Claude Mythos 5 by 3.362.559.2+3.3
Questions people ask
Which is better, Claude Mythos 5 or Claude Sonnet 5?
Claude Mythos 5 wins four of the five areas where both have results: coding, agents, reasoning and math. Claude Sonnet 5 wins none, but costs 80% less. They are level on images and charts.
Which is better for coding?
Claude Mythos 5. It wins 4 of the 4 coding tests both models report; Claude Sonnet 5 wins none.
Which is cheaper?
Claude Mythos 5 costs $10.00 per million input tokens and $50.00 per million output tokens; Claude Sonnet 5 costs $2.00 and $10.00. That makes Claude Sonnet 5 about 80% cheaper for the same work.
How do you compare the two?
We use the 28 benchmark tests both models have published scores on. The verdict counts the 9 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 19 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.