Claude Fable 5 vs DeepSeek-V4-Flash
Wins 6 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Long documents
Wins 0 of 7 areas
—
Claude Fable 5 is the stronger all-rounder.DeepSeek-V4-Flash is cheaper.
Scores updated · 49 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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 software90Claude Fable 59 of 9 tests
- ReasoningHard problems that need careful thinking60Claude Fable 56 of 6 tests
- AgentsCarrying out multi-step tasks on its own41Claude Fable 54 of 5 tests
- FactsGetting facts right instead of making them up30Claude Fable 53 of 3 tests
- MathCompetition and research-level math30Claude Fable 53 of 3 tests
- Long documentsFinding answers in very long texts10Claude Fable 51 of 1 test
- Following instructionsDoing exactly what it is asked11Even1 each
Long documents 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.
DeepSeek-V4-Flash costs 97% 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 Fable 5 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+65.8points ahead
- Short factual questions, answered correctlySimpleQA Verified+36.6points ahead
- Common-sense trick questionsSimpleBench+35.6points ahead
Where DeepSeek-V4-Flash pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+15.7points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+1.3points ahead
Every test, side by side
All 49 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5
- SWE-bench ProClaude Fable 5 by 27.48052.6+27.4
- Terminal-Bench HardClaude Fable 5 by 27.362.935.6+27.3
- LMArena · WebDevClaude Fable 5 by 195 rating points16251430+195 rating
- SWE-bench VerifiedClaude Fable 5 by 169579+16
- LiveBench · Agentic CodingClaude Fable 5 by 15.462.246.8+15.4
- SWE-bench MultilingualClaude Fable 5 by 13.386.673.3+13.3
- LiveBench · CodingClaude Fable 5 by 118675+11
- SciCodeClaude Fable 5 by 10.76150.3+10.7
- Terminal-Bench 2.1Claude Fable 5 by 5.984.678.7+5.9
AgentsClaude Fable 5
- Terminal-Bench 4.0Claude Fable 5 by 30.342.412.1+30.3
- MCP AtlasClaude Fable 5 by 15.784.769+15.7
- BrowseCompClaude Fable 5 by 14.287.473.2+14.2
- GDPValClaude Fable 5 by 8.755.646.9+8.7
- τ-Bench V3 · BankingDeepSeek-V4-Flash by 1.338.139.4+1.3
ReasoningClaude Fable 5
- SimpleBenchClaude Fable 5 by 35.681.946.3+35.6
- ARC-AGI-2Claude Fable 5 by 27.889.261.4+27.8
- Humanity's Last ExamClaude Fable 5 by 16.955.538.6+16.9
- CritPtClaude Fable 5 by 1228.616.6+12
- LiveBench · ReasoningClaude Fable 5 by 3.189.786.6+3.1
- GPQA DiamondClaude Fable 5 by 1.892.690.8+1.8
FactsClaude Fable 5
- SimpleQA VerifiedClaude Fable 5 by 36.670.734.1+36.6
- AA-Omniscience · Non-hallucinationClaude Fable 5 by 28.136.48.3+28.1
- AA-Omniscience · AccuracyClaude Fable 5 by 2565.340.4+25
MathClaude Fable 5
- FrontierMath Tier 4Claude Fable 5 by 65.890.224.4+65.8
- FrontierMath Tiers 1-3 (v2)Claude Fable 5 by 29.58757.5+29.5
- LiveBench · MathematicsClaude Fable 5 by 9.29686.8+9.2
Following instructionsEven
- IFBenchDeepSeek-V4-Flash by 15.763.579.2+15.7
- LiveBench · Instruction FollowingClaude Fable 5 by 10.375.865.5+10.3
Other results20 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.
- AA-OmniscienceClaude Fable 5 by 57.643.3-14.3+57.6
- GDPval-AA (Elo)Claude Fable 5 by 537 rating points19321395+537 rating
- GDPval-AA v2Claude Fable 5 by 534 rating points17231189+534 rating
- SWEBench Pro PublicClaude Fable 5 by 27.48052.6+27.4
- HLE (with tools)Claude Fable 5 by 18.763.845.1+18.7
- JobBenchClaude Fable 5 by 16.157.441.3+16.1
- DeepSWEClaude Fable 5 by 15.67054.4+15.6
- DeepSWE 1.1Claude Fable 5 by 15.67054.4+15.6
- AA IntelligenceClaude Fable 5 by 15.349.634.3+15.3
- livebench_languageClaude Fable 5 by 11.590.779.2+11.5
- ARC-AGI-1Claude Fable 5 by 9.598.589+9.5
- AA Agentic IndexClaude Fable 5 by 9.35141.7+9.3
- ToolathlonDeepSeek-V4-Flash by 8.661.770.3+8.6
- Toolathlon VerifiedClaude Fable 5 by 7.677.970.3+7.6
- Artificial Analysis Coding IndexClaude Fable 5 by 7.476.569.1+7.4
- τ³-Bench BankingClaude Fable 5 by 3.926.822.9+3.9
- τ²-Bench Telecom (AA run)Claude Fable 5 by 3.598.595+3.5
- StrongREJECTClaude Fable 5 by 1.398.797.4+1.3
- livebench_data_analysisClaude Fable 5 by 1.280.579.3+1.2
- Agents' Last Examtie25.725.2tie
Questions people ask
Which is better, Claude Fable 5 or DeepSeek-V4-Flash?
Claude Fable 5 wins six of the seven areas where both have results: coding, agents, reasoning, facts, math and long documents. DeepSeek-V4-Flash wins none, but costs 97% less. They are level on following instructions.
Which is better for coding?
Claude Fable 5. It wins 9 of the 9 coding tests both models report; DeepSeek-V4-Flash wins none.
Which is cheaper?
Claude Fable 5 costs $10.00 per million input tokens and $50.00 per million output tokens; DeepSeek-V4-Flash costs $0.44 and $1.32. That makes DeepSeek-V4-Flash about 97% cheaper for the same work.
How do you compare the two?
We use the 49 benchmark tests both models have published scores on. The verdict counts the 29 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 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.