Skip to main content
BenchLM
Advertise

Model comparison

GPT-5.3 Codex vs Kimi K2.5

Data verified

Head-to-head evidence from 19 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

66.69/100
Margin
7.0pts
← winning
Kimi K2.5
Moonshot AI
59.66/100
2 category wins0 category wins

Public leaderboard positions: GPT-5.3 Codex #26 (Supported); Kimi K2.5 #54 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. GPT-5.3 Codex and Kimi K2.5 share 19 comparable benchmark results. 2 of 8 categories are comparable. 2 results are unique to GPT-5.3 Codex; 44 to Kimi K2.5.

Updated July 20, 2026
Shared results
19
GPT-5.3 Codex only
2
Kimi K2.5 only
44
Comparable categories
2 / 8

Pick GPT-5.3 Codex if you want the stronger benchmark profile. Kimi K2.5 only becomes the better choice if you want the cheaper token bill or you would rather avoid the extra latency and token burn of a reasoning model.

Confidence note. This is a partial-evidence comparison with 19 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

GPT-5.3 Codex is clearly ahead on the BenchAlign aggregate, 66.69 to 59.66. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

GPT-5.3 Codex's sharpest advantage is in agentic, where it averages 71.4 against 55. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 77.3% to 50.8%.

GPT-5.3 Codex is also the more expensive model on tokens at $1.75 input / $14.00 output per 1M tokens, versus $0.60 input / $3.00 output per 1M tokens for Kimi K2.5. That is roughly 4.7x on output cost alone. GPT-5.3 Codex is the reasoning model in the pair, while Kimi K2.5 is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use. GPT-5.3 Codex gives you the larger context window at 400K, compared with 256K for Kimi K2.5.

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 scores and score margins for GPT-5.3 Codex and Kimi K2.5
CategoryGPT-5.3 CodexΔKimi K2.5
AgenticGPT-5.3 Codex71.4Margin← 16.4Kimi K2.555.0
CodingGPT-5.3 Codex67.2Margin← 7.8Kimi K2.559.4
ReasoningGPT-5.3 CodexNot measuredMarginNo overlapKimi K2.561.0
KnowledgeGPT-5.3 CodexNot measuredMarginNo overlapKimi K2.556.9
MathGPT-5.3 CodexNot measuredMarginNo overlapKimi K2.560.6
MultilingualGPT-5.3 CodexNot measuredMarginNo overlapKimi K2.582.3
MultimodalGPT-5.3 CodexNot measuredMarginNo overlapKimi K2.578.5
Inst. FollowingGPT-5.3 CodexNot measuredMarginNo overlapKimi K2.593.9

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · GPT-5.3 CodexB · Kimi K2.5

  1. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 77.3%B 50.8%
    Winner: GPT-5.3 CodexΔ 26.5
    Terminal-Bench 2.0: GPT-5.3 Codex scored 77.3%; Kimi K2.5 scored 50.8%. GPT-5.3 Codex wins this benchmark.

  2. SWE-bench Verified

    Coding
    Source ↗
    A 85%B 76.8%
    Winner: GPT-5.3 CodexΔ 8.2
    SWE-bench Verified: GPT-5.3 Codex scored 85%; Kimi K2.5 scored 76.8%. GPT-5.3 Codex wins this benchmark.

  3. SWE-bench Pro

    Coding
    Source ↗
    A 56.8%B 50.7%
    Winner: GPT-5.3 CodexΔ 6.1
    SWE-bench Pro: GPT-5.3 Codex scored 56.8%; Kimi K2.5 scored 50.7%. GPT-5.3 Codex wins this benchmark.

  4. SWE-Rebench

    Coding
    Source ↗
    A 58.2%B 58.5%
    Winner: Kimi K2.5Δ 0.3
    SWE-Rebench: GPT-5.3 Codex scored 58.2%; Kimi K2.5 scored 58.5%. Kimi K2.5 wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricGPT-5.3 CodexKimi K2.5Comparison
Input / output priceUSD per 1M tokensGPT-5.3 Codex$1.75 input / $14 outputKimi K2.5$0.6 input / $3 outputKimi K2.5 has the lower combined listed price.
Generation speedtokens per secondGPT-5.3 Codex79 tok/sKimi K2.545 tok/sGPT-5.3 Codex has the higher measured throughput.
First-answer latencyseconds to first tokenGPT-5.3 Codex88.26 sKimi K2.52.38 sKimi K2.5 reaches the first token sooner.
Context windowmaximum listed tokensGPT-5.3 Codex400KKimi K2.5256KGPT-5.3 Codex lists the larger context window.

Benchmark Deep Dive

AgenticGPT-5.3 Codex wins
BenchmarkGPT-5.3 CodexKimi K2.5Result
Terminal-Bench 2.0Source 77.3%50.8%GPT-5.3 Codex leads
OSWorld-VerifiedSource 64.7%—Not comparable
τ²-bench resultsSource 86%95.9%Kimi K2.5 leads
Gert LabsSource 57.47%45.88%GPT-5.3 Codex leads
JobBenchSource 33.7%8.7%GPT-5.3 Codex leads
BrowseCompSource —60.6%Not comparable
Claw-EvalSource —52.3%Not comparable
QwenClawBenchSource —54.3%Not comparable
τ³-bench resultsSource —65.7%Not comparable
DeepSearchQASource —77.1%Not comparable
DeepPlanningSource —14.4%Not comparable
ToolathlonSource —27.8%Not comparable
MCP AtlasSource —29.5%Not comparable
MCP-TasksSource —59.1%Not comparable
WideResearchSource —72.7%Not comparable
APEX-Agents-AASource —11.5%Not comparable
ResearchClawBenchSource —14.0%Not comparable
AA Agentic IndexSource —21.7%Not comparable
GDPval-AASource —25.4%Not comparable
GDPval-AASource —1009Not comparable
CodingGPT-5.3 Codex wins
BenchmarkGPT-5.3 CodexKimi K2.5Result
SWE-bench VerifiedSource 85%76.8%GPT-5.3 Codex leads
SWE-bench ProSource 56.8%50.7%GPT-5.3 Codex leads
SWE-RebenchSource 58.2%58.5%Kimi K2.5 leads
Vibe Code BenchSource 61.77%—Not comparable
AA-SciCodeSource 53.2%49.0%GPT-5.3 Codex leads
SWE-bench Verified*Source —70.8%Not comparable
LiveCodeBench v6Source —85.0%Not comparable
SWE MultilingualSource —73%Not comparable
React Native EvalsSource —77.2%Not comparable
SciCodeSource —48.7%Not comparable
AA Coding IndexSource —46.8%Not comparable
Reasoning
BenchmarkGPT-5.3 CodexKimi K2.5Result
AA-LCRSource 74.0%65.3%GPT-5.3 Codex leads
CritPtSource 16.9%3.1%GPT-5.3 Codex leads
LongBench v2Source —61%Not comparable
Knowledge
BenchmarkGPT-5.3 CodexKimi K2.5Result
Artificial Analysis Intelligence IndexSource 44.3%35.4%GPT-5.3 Codex leads
AA-GPQA DiamondSource 91.5%87.9%GPT-5.3 Codex leads
AA-HLESource 39.9%29.4%GPT-5.3 Codex leads
AA-Omniscience IndexSource 9.9%-8.1%GPT-5.3 Codex leads
AA-Omniscience AccuracySource 51.8%34.3%GPT-5.3 Codex leads
AA-Omniscience Hallucination RateSource 86.9%64.6%Kimi K2.5 leads
GPQASource —87.6%Not comparable
GPQA-DSource —87.6%Not comparable
SuperGPQASource —69.2%Not comparable
MMLU-ProSource —87.1%Not comparable
MMLU-Pro (Arcee)Source —87.1%Not comparable
HLESource —30.1%Not comparable
Math
BenchmarkGPT-5.3 CodexKimi K2.5Result
AIME 2025Source —96.1%Not comparable
AIME26Source —95.8%Not comparable
AIME25 (Arcee)Source —96.3%Not comparable
HMMT Feb 2025Source —95.4%Not comparable
HMMT Nov 2025Source —91.1%Not comparable
HMMT Feb 2026Source —87.1%Not comparable
MMAnswerBenchSource —81.8%Not comparable
FrontierMath v2 (Tiers 1-3)Source —27.900%Not comparable
FrontierMath v2 (Tier 4)Source —4.200%Not comparable
Multilingual
BenchmarkGPT-5.3 CodexKimi K2.5Result
MMLU-ProXSource —82.3%Not comparable
NOVA-63Source —56.0%Not comparable
Multimodal
BenchmarkGPT-5.3 CodexKimi K2.5Result
AA-MMMU-ProSource 78.5%75.4%GPT-5.3 Codex leads
Design Arena WebsiteSource 11951282Kimi K2.5 leads
MMMU-ProSource —78.5%Not comparable
Video-MMESource —87.4%Not comparable
MMVUSource —80.4%Not comparable
VideoMMMUSource —86.6%Not comparable
Inst. Following
BenchmarkGPT-5.3 CodexKimi K2.5Result
AA-IFBenchSource 75.4%70.2%GPT-5.3 Codex leads
IFEvalSource —93.9%Not comparable
Frequently Asked Questions (3)

Which is better, GPT-5.3 Codex or Kimi K2.5?

GPT-5.3 Codex is ahead on BenchLM's BenchAlign leaderboard, 66.69 to 59.66. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 77.3% and 50.8%.

Which is better for coding, GPT-5.3 Codex or Kimi K2.5?

GPT-5.3 Codex has the edge for coding in this comparison, averaging 67.2 versus 59.4. Inside this category, SWE-bench Verified is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, GPT-5.3 Codex or Kimi K2.5?

GPT-5.3 Codex has the edge for agentic tasks in this comparison, averaging 71.4 versus 55. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.

Self-host vs API cost

Estimates at 50,000 req/day · 1000 tokens/req average.

GPT-5.3 Codex
API / mo$11,813
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Kimi K2.5
API / mo$2,700
Self-host / mo$5,221
Break-even132M/day
Model the full break-even

Related Comparisons

Explore More

OpenAI Compare Pricing Methodology Find Your Best LLM Overall Rankings
Last updated: July 20, 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.