Enter opens the highlighted result. Esc closes.

Claude Haiku 4.5 vs GPT-6 Sol

GPT-6 Sol leads Claude Haiku 4.5 on coding, 76.0 to 35.6 on the Vibe Coding Index, and Claude Haiku 4.5 costs 2.0x less per output token ($5.00 vs $10.00 per million). Updated October 22, 2026; 69 benchmarks have a result for both.

Replace Claude Haiku 4.5 Replace GPT-6 Sol
Claude Haiku 4.5Anthropic
Coding
35.6
Overall
41.0
Output $/M
$5.00
17
GPT-6 SolOpenAI
Coding
76.0
Overall
74.6
Output $/M
$10.00
52
vs boards won of 69 shared

Claude Haiku 4.5 or GPT-6 Sol: which to pick

Pick GPT-6 Sol for coding: it scores 76.0 on the Vibe Coding Index against 35.6 for Claude Haiku 4.5.

Pick Claude Haiku 4.5 for output tokens 2.0x cheaper ($5.00 vs $10.00 per million) and work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.

Model AClaude Haiku 4.5Anthropic35.6/1004 independent benchmarks$0.56 per solved taskModel BGPT-6 SolOpenAI76.0/1005 independent benchmarksNo cost per solved task
Shared results
69
Claude Haiku 4.5 only
26
GPT-6 Sol only
32
Like-for-like categories
8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Claude Haiku 4.5 35.6 from 4 benchmarks, GPT-6 Sol 76.0 from 5 benchmarks.
Widest category
Agentic work, 39.6 points apart on average over 8 shared benchmarks.
Widest single gap
Vending-Bench 2: Claude Haiku 4.5 $459 against GPT-6 Sol $14,428, 87.3 points apart on a 0 to 100 scale.
Output price
Claude Haiku 4.5 $5.00 against GPT-6 Sol $10.00 per million output tokens, 2.0x apart.
Context window
Claude Haiku 4.5 takes 200K tokens, GPT-6 Sol 1.05M.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryClaude Haiku 4.5BarsGPT-6 SolTies
Coding28
Intelligence02
Agentic work08
Reasoning26
Math12
Knowledge42
Web design09
General815
All shared boards1752
CodingLike-for-likeSix of these boards feed the Vibe Coding Index28
Coding benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
SWE-bench Verified66.6%Not measuredOnly Claude Haiku 4.5 measured
SWE-bench Multilingual64.7%Not measuredOnly Claude Haiku 4.5 measured
LiveBench coding48.1%93.1%GPT-6 Sol
LMArena WebDev13301683GPT-6 Sol
AA Coding Index43.9Not measuredOnly Claude Haiku 4.5 measured
LiveCodeBench (Vals AI)41.2%Not measuredOnly Claude Haiku 4.5 measured
SWE-bench Verified (Vals AI)66.6%Not measuredOnly Claude Haiku 4.5 measured
Terminal-Bench 4 (Vals AI)Not measured44.4%Only GPT-6 Sol measured
Vibe Code Bench (Vals AI)11.4%87.8%GPT-6 Sol
Code migration (Vals AI)7.5%57.2%GPT-6 Sol
ProgramBench (Vals AI)0.0%2.0%GPT-6 Sol
IOI (Vals AI)Not measured82.6%Only GPT-6 Sol measured
Terminal-Bench 2.1 (Vals AI)43.8%83.2%GPT-6 Sol
CyberBench (Vals AI)Not measured78.0%Only GPT-6 Sol measured
SciCodeNot measured57.6%Only GPT-6 Sol measured
Terminal-Bench35.5%Not measuredOnly Claude Haiku 4.5 measured
FrontierCodeNot measured49.3%Only GPT-6 Sol measured
WeirdML45.4%Not measuredOnly Claude Haiku 4.5 measured
ALE-BenchNot measured2462.0Only GPT-6 Sol measured
LMArena Text: coding14541448Claude Haiku 4.5
LMArena Text: software and IT14391433Claude Haiku 4.5
LMArena WebDev: React13261685GPT-6 Sol
LMArena WebDev: HTML13181685GPT-6 Sol
LMArena WebDev: from an imageNot measured1660Only GPT-6 Sol measured
Neither model has a result on the other 15 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
AA Intelligence Index16.947.6GPT-6 Sol
Epoch Capabilities Index142.2163.7GPT-6 Sol
Agentic workLike-for-likePublic boards, shown as published, kept out of the index08
Agentic work benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
AA Agentic Index8.0Not measuredOnly Claude Haiku 4.5 measured
Design Arena: agents building web appsNot measured1231Only GPT-6 Sol measured
Design Arena: agents building full-stack appsNot measured1238Only GPT-6 Sol measured
Design Arena: agents building mobile appsNot measured1217Only GPT-6 Sol measured
LiveBench research42.2%69.3%GPT-6 Sol
LiveBench browser40.0%92.0%GPT-6 Sol
LiveBench long agentic11.1%66.9%GPT-6 Sol
SREBench (Vals AI)Not measured30.5%Only GPT-6 Sol measured
Finance Agent (Vals AI)31.0%49.0%GPT-6 Sol
Tax Agent Bench (Vals AI)5.9%15.0%GPT-6 Sol
Legal Research Bench (Vals AI)10.6%28.9%GPT-6 Sol
Excel Modeling Benchmark (Vals AI)22.7%71.5%GPT-6 Sol
BioMysteryBench (Vals AI)Not measured74.8%Only GPT-6 Sol measured
Terminal-Bench-Science (Vals AI)Not measured30.0%Only GPT-6 Sol measured
Vals Science IndexNot measured45.0%Only GPT-6 Sol measured
APEX-AgentsNot measured54.3%Only GPT-6 Sol measured
BALROG31.2%Not measuredOnly Claude Haiku 4.5 measured
DeepResearch Bench45.5%Not measuredOnly Claude Haiku 4.5 measured
SpireClimbNot measured7.5%Only GPT-6 Sol measured
Vending-Bench 2$459$14,428GPT-6 Sol
Berkeley Function Calling Leaderboard (BFCL)68.7%Not measuredOnly Claude Haiku 4.5 measured
Neither model has a result on the other 20 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index26
Reasoning benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
GPQA Diamond72.2%Not measuredOnly Claude Haiku 4.5 measured
LiveBench reasoning61.7%88.7%GPT-6 Sol
MysteryMechanism (Vals AI)Not measured30.2%Only GPT-6 Sol measured
MMMU Pro (Vals AI)46.1%Not measuredOnly Claude Haiku 4.5 measured
Humanity's Last Exam (Artificial Analysis run)Not measured47.9%Only GPT-6 Sol measured
GPQA Diamond (Epoch AI run)71.2%94.3%GPT-6 Sol
ARC-AGI47.7%95.5%GPT-6 Sol
ARC-AGI-24.0%89.6%GPT-6 Sol
Chess puzzles8.0%56.0%GPT-6 Sol
Mystery game puzzlesNot measured56.0%Only GPT-6 Sol measured
EBR-benchNot measured53.3%Only GPT-6 Sol measured
LMArena Text: hard prompts14211422GPT-6 Sol
LMArena Text: expert14431437Claude Haiku 4.5
LMArena Text: hard prompts in English14351420Claude Haiku 4.5
LMArena Vision: diagramsNot measured1270Only GPT-6 Sol measured
LMArena Vision: homeworkNot measured1286Only GPT-6 Sol measured
Neither model has a result on the other 6 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index12
Math benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
AIME82.7%Not measuredOnly Claude Haiku 4.5 measured
MGSM (Vals AI)92.2%Not measuredOnly Claude Haiku 4.5 measured
OTIS Mock AIME66.7%100.0%GPT-6 Sol
FrontierMath, tiers 1 to 3Not measured89.8%Only GPT-6 Sol measured
FrontierMath, tier 4Not measured90.0%Only GPT-6 Sol measured
ProofBenchNot measured83.0%Only GPT-6 Sol measured
MATH level 596.4%Not measuredOnly Claude Haiku 4.5 measured
LMArena Text: math13981407GPT-6 Sol
LMArena Text: mathematical work14301424Claude Haiku 4.5
Neither model has a result on the other 4 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index42
Knowledge benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
MMLU Pro78.7%Not measuredOnly Claude Haiku 4.5 measured
LegalBench (Vals AI)81.2%Not measuredOnly Claude Haiku 4.5 measured
CaseLaw (Vals AI)56.5%Not measuredOnly Claude Haiku 4.5 measured
CorpFin (Vals AI)60.6%Not measuredOnly Claude Haiku 4.5 measured
TaxEval (Vals AI)67.5%Not measuredOnly Claude Haiku 4.5 measured
MortgageTax (Vals AI)62.2%Not measuredOnly Claude Haiku 4.5 measured
MedQA (Vals AI)79.6%Not measuredOnly Claude Haiku 4.5 measured
MedCode (Vals AI)32.7%47.1%GPT-6 Sol
SimpleQA Verified13.2%60.7%GPT-6 Sol
LMArena Text: science14061389Claude Haiku 4.5
LMArena Text: medicine and healthcare13941366Claude Haiku 4.5
LMArena Text: legal and government14001396Claude Haiku 4.5
LMArena Text: business and finance14001380Claude Haiku 4.5
FACTS Benchmark SuiteNot measured72.5%Only GPT-6 Sol measured
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index09
Web design benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
Design Arena: websites11221292GPT-6 Sol
Design Arena: UI components10971290GPT-6 Sol
Design Arena: game dev10971311GPT-6 Sol
Design Arena: data viz11241269GPT-6 Sol
Design Arena: 3D10711314GPT-6 Sol
Design Arena: all code categories11141300GPT-6 Sol
Design Arena: SVG10251315GPT-6 Sol
Design Arena: ASCII art11401222GPT-6 Sol
LiveBench design59.9%82.2%GPT-6 Sol
GeneralLike-for-likePublic boards, shown as published, kept out of the index815
General benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
LiveBench data analysis59.3%81.2%GPT-6 Sol
Public Benefits Bench (Vals AI)54.3%56.6%GPT-6 Sol
SAGE (Vals AI)31.8%44.8%GPT-6 Sol
MedScribe (Vals AI)85.2%82.0%Claude Haiku 4.5
Vals Multimodal Index42.9%Not measuredOnly Claude Haiku 4.5 measured
Long context reasoning (AA-LCR)Not measured83.7%Only GPT-6 Sol measured
DTBench73.6%97.3%GPT-6 Sol
LMCA30.9%59.1%GPT-6 Sol
Furniture assemblyNot measured58.3%Only GPT-6 Sol measured
ExploitBench13.7%Not measuredOnly Claude Haiku 4.5 measured
LMArena Text13971398GPT-6 Sol
LMArena Text: instruction following14141413Claude Haiku 4.5
LMArena Text: creative writing13711378GPT-6 Sol
LMArena VisionNot measured1240Only GPT-6 Sol measured
LMArena Text: English14141396Claude Haiku 4.5
LMArena Text: non-English13771388GPT-6 Sol
LMArena Text: Chinese14181412Claude Haiku 4.5
LMArena Text: German13741379GPT-6 Sol
LMArena Text: Japanese13381377GPT-6 Sol
LMArena Text: Korean13461336Claude Haiku 4.5
LMArena Text: French14111399Claude Haiku 4.5
LMArena Text: Spanish14211387Claude Haiku 4.5
LMArena Text: Russian13811402GPT-6 Sol
LMArena Text: Polish13431372GPT-6 Sol
LMArena Text: multi-turn14091414GPT-6 Sol
LMArena Text: longer queries14271416Claude Haiku 4.5
LMArena Text: writing and language13861389GPT-6 Sol
LMArena Text: entertainment and media13691373GPT-6 Sol
LMArena Vision: EnglishNot measured1229Only GPT-6 Sol measured
LMArena Vision: reading textNot measured1258Only GPT-6 Sol measured
LMArena Vision: creative writingNot measured1219Only GPT-6 Sol measured
LMArena Document1420Not measuredOnly Claude Haiku 4.5 measured
Neither model has a result on the other 14 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Claude Haiku 4.5 and GPT-6 Sol
BenchmarkClaude Haiku 4.5GPT-6 SolLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Claude Haiku 4.5$0.0035Fits in one request
GPT-6 Sol$0.0070Fits in one request

Claude Haiku 4.5 costs less

Repository review

50K input + 3K output tokens

Claude Haiku 4.5$0.065Fits in one request
GPT-6 Sol$0.130Fits in one request

Claude Haiku 4.5 costs less

Agent loop

200K input + 10K output tokens

Claude Haiku 4.5$0.250Exceeds the context window
GPT-6 Sol$0.500Fits in one request

Claude Haiku 4.5 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Claude Haiku 4.5 and GPT-6 Sol
SpecificationAnthropic Claude Haiku 4.5OpenAI GPT-6 Sol
MakerAnthropicOpenAI
WeightsNo public weights listedNo public weights listed
ReleasedOctober 15, 2025September 22, 2026
Context window200K1.05M
Input $/Mtok$1.00$2.00
Output $/Mtok$5.00$10.00
Coding index35.676.0
Independent measurements behind it45
Cost per solved task$0.56Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Claude Haiku 4.5 better than GPT-6 Sol for coding?

Pick GPT-6 Sol for coding: it scores 76.0 on the Vibe Coding Index against 35.6 for Claude Haiku 4.5. Pick Claude Haiku 4.5 for output tokens 2.0x cheaper ($5.00 vs $10.00 per million) and work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.

Which is cheaper, Claude Haiku 4.5 or GPT-6 Sol?

Claude Haiku 4.5 costs $1.00 per million input tokens and $5.00 per million output tokens; GPT-6 Sol costs $2.00 and $10.00. Cost per solved task is known for Claude Haiku 4.5 only.

Which has the bigger context window, Claude Haiku 4.5 or GPT-6 Sol?

GPT-6 Sol, with 1.05M tokens against 200K.

How many benchmarks is the Claude Haiku 4.5 vs GPT-6 Sol comparison based on?

69 benchmarks have a result for both models as of October 22, 2026. Claude Haiku 4.5's coding index rests on 4 independent benchmarks and GPT-6 Sol's on 5. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/claude-haiku-4-5-vs-gpt-6-sol" title="Claude Haiku 4.5 vs GPT-6 Sol, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/claude-haiku-4-5-vs-gpt-6-sol">LLMBenchmark.ai, Claude Haiku 4.5 vs GPT-6 Sol</a></p>