Model comparison
Claude Fable 5 vs DeepSeek V4 Pro (Max)
Head-to-head evidence from 26 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: Claude Fable 5 #3 (Supported); DeepSeek V4 Pro (Max) #57 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Claude Fable 5 and DeepSeek V4 Pro (Max) share 26 comparable benchmark results. 2 of 8 categories are comparable. 8 results are unique to Claude Fable 5; 22 to DeepSeek V4 Pro (Max).
Updated July 24, 2026- Shared results
- 26
- Claude Fable 5 only
- 8
- DeepSeek V4 Pro (Max) only
- 22
- Comparable categories
- 2 / 8
Pick Claude Fable 5 if you want the stronger benchmark profile. DeepSeek V4 Pro (Max) only becomes the better choice if you want the cheaper token bill.
Confidence note. This is a partial-evidence comparison with 26 shared benchmark results across 6 evidence categories; 2 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
Claude Fable 5 is clearly ahead on the BenchAlign aggregate, 82.76 to 59.09. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Claude Fable 5's sharpest advantage is in coding, where it averages 89.2 against 70.9. The single biggest benchmark swing on the page is SWE-bench Pro, 80% to 55.4%.
Claude Fable 5 is also the more expensive model on tokens at $10.00 input / $50.00 output per 1M tokens, versus $0.43 input / $0.87 output per 1M tokens for DeepSeek V4 Pro (Max). That is roughly 57.5x on output cost alone. Claude Fable 5 gives you the larger context window at 1M+, compared with 1M for DeepSeek V4 Pro (Max).
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | Claude Fable 5 | Δ | DeepSeek V4 Pro (Max) |
|---|---|---|---|
| Coding | Claude Fable 589.2 | Margin← 18.3 | DeepSeek V4 Pro (Max)70.9 |
| Agentic | Claude Fable 584.6 | Margin← 10.1 | DeepSeek V4 Pro (Max)74.5 |
| Knowledge | Claude Fable 5Not measured | MarginNo overlap | DeepSeek V4 Pro (Max)60.1 |
| Math | Claude Fable 5Not measured | MarginNo overlap | DeepSeek V4 Pro (Max)95.2 |
| Multimodal | Claude Fable 557.9 | MarginNo overlap | DeepSeek V4 Pro (Max)Not measured |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
SWE-bench Pro
CodingA 80%B 55.4%Winner: Claude Fable 5Δ 24.6SWE-bench Pro: Claude Fable 5 scored 80%; DeepSeek V4 Pro (Max) scored 55.4%. Claude Fable 5 wins this benchmark. - Source ↗
Terminal-Bench 2.0
AgenticA 84.3%B 67.9%Winner: Claude Fable 5Δ 16.4Terminal-Bench 2.0: Claude Fable 5 scored 84.3%; DeepSeek V4 Pro (Max) scored 67.9%. Claude Fable 5 wins this benchmark. - Source ↗
SWE-bench Verified
CodingA 95%B 80.6%Winner: Claude Fable 5Δ 14.4SWE-bench Verified: Claude Fable 5 scored 95%; DeepSeek V4 Pro (Max) scored 80.6%. Claude Fable 5 wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Claude Fable 5 | DeepSeek V4 Pro (Max) | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Claude Fable 5$10 input / $50 output | DeepSeek V4 Pro (Max)$0.435 input / $0.87 output | DeepSeek V4 Pro (Max) has the lower combined listed price. |
| Generation speedtokens per second | Claude Fable 5Not available | DeepSeek V4 Pro (Max)Not available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | Claude Fable 5Not available | DeepSeek V4 Pro (Max)Not available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | Claude Fable 51M+ | DeepSeek V4 Pro (Max)1M | Claude Fable 5 lists the larger context window. |
Benchmark Deep Dive
AgenticClaude Fable 5 wins19 benchmarks
| Benchmark | Claude Fable 5 | DeepSeek V4 Pro (Max) | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 84.3% | 67.9% | Claude Fable 5 leads |
| OSWorld-VerifiedSource | 85% | — | Not comparable |
| GDPval-AASource | 1747 | 1306 | Claude Fable 5 leads |
| AA Agentic IndexSource | 52.8% | 36.4% | Claude Fable 5 leads |
| τ²-bench resultsSource | 98.5% | 96.2% | Claude Fable 5 leads |
| GDPval-AASource | 62.3% | 40.3% | Claude Fable 5 leads |
| AA BriefcaseSource | 1574 | 931 | Claude Fable 5 leads |
| AA AutomationBenchSource | 48.6% | — | Not comparable |
| AA EnterpriseOps-GymSource | 51.1% | 40.4% | Claude Fable 5 leads |
| AA Harvey LABSource | 93.6% | 84.4% | Claude Fable 5 leads |
| AA Tau3 BankingSource | 26.8% | 25.8% | Claude Fable 5 leads |
| terminalBenchHardSource | 62.9% | 46.2% | Claude Fable 5 leads |
| aaTerminalBench21Source | 84.6% | 64% | Claude Fable 5 leads |
| BrowseCompSource | — | 83.4% | Not comparable |
| HLE w/ toolsSource | — | 48.2% | Not comparable |
| MCP AtlasSource | — | 73.6% | Not comparable |
| ToolathlonSource | — | 51.8% | Not comparable |
| APEX-Agents-AASource | — | 24.3% | Not comparable |
| AA ITBenchSource | — | 38.3% | Not comparable |
CodingClaude Fable 5 wins12 benchmarks
| Benchmark | Claude Fable 5 | DeepSeek V4 Pro (Max) | Result |
|---|---|---|---|
| SWE-bench VerifiedSource | 95% | 80.6% | Claude Fable 5 leads |
| SWE-bench ProSource | 80% | 55.4% | Claude Fable 5 leads |
| FrontierCode 1.1 MainSource | 53.5% | — | Not comparable |
| Terminal-Bench 2.0Source | 84.3% | 67.9% | Claude Fable 5 leads |
| cursorBench31Source | 70.6% | — | Not comparable |
| cursorBench32Source | 70.5% | — | Not comparable |
| VulcanBench v3Source | 87.0% | — | Not comparable |
| AA Coding IndexSource | 76.5% | 59.4% | Claude Fable 5 leads |
| AA-SciCodeSource | 60.2% | 50.0% | Claude Fable 5 leads |
| CodeforcesSource | — | 3206.0 | Not comparable |
| SWE MultilingualSource | — | 76.2% | Not comparable |
| Vibe Code BenchSource | — | 49.93% | Not comparable |
Reasoning4 benchmarks
Knowledge13 benchmarks
| Benchmark | Claude Fable 5 | DeepSeek V4 Pro (Max) | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 59.9% | 44.3% | Claude Fable 5 leads |
| AA-GPQA DiamondSource | 92.6% | 88.8% | Claude Fable 5 leads |
| AA-HLESource | 53.3% | 35.9% | Claude Fable 5 leads |
| AA-Omniscience IndexSource | 40.2% | -10.0% | Claude Fable 5 leads |
| AA-Omniscience AccuracySource | 61.4% | 43.3% | Claude Fable 5 leads |
| AA-Omniscience Hallucination RateSource | 54.9% | 94.0% | Claude Fable 5 leads |
| MMLU-ProSource | — | 87.5% | Not comparable |
| SimpleQASource | — | 57.9% | Not comparable |
| Chinese-SimpleQASource | — | 84.4% | Not comparable |
| GPQASource | — | 90.1% | Not comparable |
| GPQA-DSource | — | 90.1% | Not comparable |
| HLESource | — | 37.7% | Not comparable |
| AA Openness IndexSource | — | 50.0% | Not comparable |
Math4 benchmarks
Multimodal3 benchmarks
Inst. Following1 benchmarks
| Benchmark | Claude Fable 5 | DeepSeek V4 Pro (Max) | Result |
|---|---|---|---|
| AA-IFBenchSource | 63.5% | 76.5% | DeepSeek V4 Pro (Max) leads |
Frequently Asked Questions (3)
Which is better, Claude Fable 5 or DeepSeek V4 Pro (Max)?
Claude Fable 5 is ahead on BenchLM's BenchAlign leaderboard, 82.76 to 59.09. The biggest single separator in this matchup is SWE-bench Pro, where the scores are 80% and 55.4%.
Which is better for coding, Claude Fable 5 or DeepSeek V4 Pro (Max)?
Claude Fable 5 has the edge for coding in this comparison, averaging 89.2 versus 70.9. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, Claude Fable 5 or DeepSeek V4 Pro (Max)?
Claude Fable 5 has the edge for agentic tasks in this comparison, averaging 84.6 versus 74.5. Inside this category, AA Briefcase is the benchmark that creates the most daylight between them.