VECTOR WIREAI INTELLIGENCE
NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%
PKT
SEEDRefresh Models Deals Regulatory Sources

Claude Opus 4.8 vs DeepSeek-V4-Flash

AnthropicvsDeepSeek55 shared benchmarks469 head-to-head
BenchmarkClaude Opus 4.8DeepSeek-V4-Flash
AA Agentic Index49.461.3
AA Intelligence57.351.8
AA-LCR7374.3
AA-Omniscience28.8-14.3
Agents' Last Exam2725.2
aime_202610095.8
arc_agi_192.584
ARC-AGI-272.146
arena_elo14821439
arena_text_factuality14641434
Artificial Analysis Coding Index74.369.1
AutomationBench (Public)27.225.1
AutomationBench Public27.225.1
browsecomp88.573.2
coding_arena_elo15391431
critpt20.916.6
cybergym83.176.7
DeepSWE5954.4
DSBench-FullStack71.668.7
DSBench-Hard71.759.6
gdpval54.252.9
GDPval-AA (Elo)18901395
GDPval-AA v215931189
GPQA Diamond93.690.8
HLE57.938.6
HLE (with tools)57.945.1
HLE (wo / w tools)57.951.5
hmmt_feb_202696.794.8
IFBench62.279.2
imo_answer_bench83.588.4
livebench_agentic_coding50.546.8
livebench_coding81.875
livebench_data_analysis6679.3
livebench_instruction_following7265.5
livebench_language79.779.2
livebench_math94.386.8
livebench_reasoning89.286.6
MCP Atlas83.669
nl2repo69.754.2
OmniScience Accuracy48.840.4
OmniScience Non-Hallucination60.715.6
scicode53.549.9
simplebench64.846.3
simpleqa_verified39.545
SWE-bench Multilingual84.473.3
SWE-bench Pro69.252.6
SWE-bench Verified88.679
Tau 3 Banking27.622.9
TauBench V3 - Banking34.239.4
Terminal-Bench 2.074.656.9
Terminal-Bench 2.18582.7
Terminal-Bench Hard58.338.6
toolathlon59.970.3
τ²-Bench Telecom (AA run)94.495.6
τ³-Bench27.622.9

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.