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-V3.2

AnthropicvsDeepSeek43 shared benchmarks421 head-to-head
BenchmarkClaude Opus 4.8DeepSeek-V3.2
AA Agentic Index49.439.8
AA Intelligence57.332.8
AA-LCR7370.7
AA-Omniscience28.8-22.5
aime_202610095.1
arc_agi_192.557
ARC-AGI-272.14
arena_text_factuality14641433
Artificial Analysis Coding Index74.344.2
browsecomp88.551.4
coding_arena_elo15391368
critpt20.92.9
cybergym83.117.3
deepsearchqa_f193.160.9
frontiermath_tier_431.32.1
gdpval54.218.8
GPQA Diamond93.684
HealthBench59.346.9
HLE57.940.8
HLE (with tools)57.940.8
hmmt_feb_202696.784.1
hmmt_nov_202596.590.2
IFBench62.261
imo_answer_bench83.578.3
longbench_v269.159.8
MCP Atlas83.662.2
OmniScience Accuracy48.833
OmniScience Non-Hallucination60.717.3
PaperBench80.347.1
ResearchRubrics73.555.8
scicode53.539
simpleqa_verified39.527.5
SWE-bench Multilingual84.470.2
SWE-bench Pro69.215.6
SWE-bench Verified88.673.1
TauBench V3 - Banking34.218.8
Terminal-Bench 2.074.646.4
Terminal-Bench 2.18546.8
Terminal-Bench Hard58.335.6
tool_decathlon59.935.2
toolathlon59.935.2
τ²-Bench Telecom (AA run)94.490.6
τ³-Bench27.669.2

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.