P99lab vs Anthropic vs DeepSeek vs Google

Head to head on 22 boards: Anthropic places best on 15, Google places best on 5, P99lab places best on 1 and DeepSeek places best on 1. All 4 ranked on 0 of 28 boards.

Two at a time on this screen. Tap one to swap it in.

PP99labAnthropicDeepSeekGoogle

Board	Read	PP99lab	Anthropic	DeepSeek	Google
Chat2
Text ArenaArena rating1dyesterday	Oct 261dyesterday		#21,509Claude Opus 5.5	#291,476DeepSeek V4.1 Flash	#11,517best of the pickedGemini 4 Argon
MultiChallengeScore1dyesterday	Oct 261dyesterday		#867.7%Claude Haiku 5.5	#2846.1%DeepSeek V3P1	#278.7%best of the pickedGemini 4 Argon
Reasoning7
AA IntelligenceIndex1dyesterday	Oct 261dyesterday		#157.6best of the pickedClaude Opus 5.5	#1445.1DeepSeek V4.1 Flash	#752.6Gemini 4 Argon
GPQA DiamondAccuracy1dyesterday	Oct 261dyesterday		#395.6%Claude Sonnet 5.5	#1993.2%DeepSeek V4.1 Flash	#196.4%best of the pickedGemini 4 Argon
Epoch ECIECI1dyesterday	Oct 261dyesterday		#1168.8best of the pickedClaude Opus 5.5	#32155.9DeepSeek V4.1 Flash	#5166.4Gemini 4 Argon
LiveBenchGlobal average1dyesterday	Oct 261dyesterday		#183.4%best of the pickedClaude Fable 5.1	#881.1%DeepSeek V4.1 Flash	#1578.8%Gemini 3.7 Flash
MMLU-ProAccuracy · stale1dyesterday	Oct 261dyesterday		#489.1%Claude Opus 4.6	#3385.0%DeepSeek-V3.2	#191.2%best of the pickedGemini 3.1 Pro Preview
ARC-AGI-2Score1dyesterday	Oct 261dyesterday		#393.3%best of the pickedClaude Opus 5.5	#1878.1%DeepSeek V4.1 Flash	#591.3%Gemini 4 Argon
HLEAccuracy1dyesterday	Oct 261dyesterday		#246.5%best of the pickedClaude Fable 5.1		#346.4%Gemini 3.1 Pro Preview
Coding4
WebDev ArenaArena rating1dyesterday	Oct 261dyesterday		#11,807best of the pickedClaude Opus 5.5	#121,665DeepSeek V4.1 Flash	#101,674Gemini 4 Argon
SWE-bench VerifiedResolved · stale1dyesterday	Oct 261dyesterday		#176.8%best of the pickedClaude Opus 4.5	#1170.0%DeepSeek-V3.2	#275.8%Gemini 3 Flash
Aider PolyglotPass rate · stale1dyesterday	Oct 261dyesterday		#1072.0%Claude Opus 4	#874.2%DeepSeek-V3.2	#383.1%best of the pickedGemini 2.5 Pro (Jun 2025)
SWE-Bench ProResolved1dyesterday	Oct 261dyesterday		#198.0%best of the pickedClaude Opus 5		#392.2%Gemini 4 Argon
Agents and tools3
BFCLOverall accuracy · stale1dyesterday	Oct 261dyesterday		#177.5%best of the pickedClaude Opus 4.5	#1256.7%DeepSeek-V3.2	#372.5%Gemini 3 Pro
MCP AtlasPass rate1dyesterday	Oct 261dyesterday		#287.2%best of the pickedClaude Fable 5.1		#683.6%Gemini 3.5 Flash
Search ArenaArena rating1dyesterday	Oct 261dyesterday		#41,233best of the pickedClaude Fable 5.1		#141,207Gemini 3.1 Pro Grounding
Vision2
Vision ArenaArena rating1dyesterday	Oct 261dyesterday		#11,309best of the pickedClaude Fable 5	#391,261DeepSeek V4.1 Flash	#51,295Gemini 3.7 Flash
Document ArenaArena rating1dyesterday	Oct 261dyesterday		#11,499best of the pickedClaude Fable 5		#211,461Gemini 3.5 Flash
Image2
Image EditArena rating1dyesterday	Oct 261dyesterday				#51,426Gemini Nano Banana 2.1
Text-to-ImageArena rating1dyesterday	Oct 261dyesterday				#61,328Gemini Nano Banana 2.1
Video2
Text-to-VideoArena rating1dyesterday	Oct 261dyesterday				#11,516Gemini Omni 1.1 Flash
Image-to-VideoArena rating1dyesterday	Oct 261dyesterday				#21,488Gemini Omni 1.1 Flash
Speech2
Open ASRAverage WER1dyesterday	Oct 261dyesterday	#34.98 WERbest of the pickedaudrey-0.7 (p99lab)			#5611.86 WERgemma-4-E4B-it (google)
Speech ArenaArena rating1dyesterday	Oct 261dyesterday				#51,280Gemini 3.8 Flash TTS
Embeddings1
MTEBMean task score1dyesterday	Oct 261dyesterday				#568.4%gemini-embedding-001 (google)
Speed1
AA SpeedOutput speed1dyesterday	Oct 261dyesterday		#3234 tok/sbest of the pickedClaude Haiku 5.5	#8181 tok/sDeepSeek V4 Flash Vision	#11132 tok/sGemini 3.8 Flash
Price1
AA PriceBlended price1dyesterday	Oct 261dyesterday		#3$0.20best of the pickedClaude Haiku 5.5	#9$0.53DeepSeek V4.1 Flash	#15$1.50Gemini 3.8 Flash
Usage1
OpenRouterWeekly tokens1dyesterday	Oct 261dyesterday		#68.3T tokClaude Opus 5.5	#133.6T tokbest of the pickedDeepSeek V4.1 Flash

Green marks the best placing among the picked where two or more are ranked; ties are marked for each. Ranks are each board's own order (lower is better), scores as the board published them. Dash: not on that board.