Coding work
Code generation, repair, and software-engineering tasks
Gemini 2.5 Pro
Gemini 2.5 Pro leads on the public coding lane, 24.4 to 19.6, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited
Claude Haiku 4.5 vs Gemini 2.5 Pro
Updated September 29, 2026. Rank says Gemini 2.5 Pro is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Both of these models will change. Get the price, version and retirement notices for the pair, each with its source. Follow model changes
Share or export
Decision reading
Gemini 2.5 Pro has the higher public score estimate, 50.17 versus 42.47, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 4 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
- Shared results
- 4
- Claude Haiku 4.5 only
- 6
- Gemini 2.5 Pro only
- 4
- Like-for-like categories
- 1 / 8
Estimated: Claude Haiku 4.5 · Supported: Gemini 2.5 ProHow the comparison works
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Long documents
Prompts that approach the documented context limit
Gemini 2.5 Pro
Gemini 2.5 Pro has the larger documented context window.
Confidence: documentedChat turn cost
1K fresh input + 500 output tokens
Claude Haiku 4.5
Claude Haiku 4.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Show secondary and unsupported calls
Repository review cost
50K fresh input + 3K output tokens
Claude Haiku 4.5
Claude Haiku 4.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-ratesAgentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Gemini 2.5 Pro is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
Confidence: limitedCache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
The page does not recommend a cost winner because at least one model cannot fit the stated workload in one request. Claude Haiku 4.5 does not fit this workload in one request.
Confidence: listed-rates
Which one for a specific job
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
19.6Claude Haiku 4.524.4Gemini 2.5 Pro
Like-for-like · BenchAlign v5.7
Gemini 2.5 Pro leads the like-for-like coding row, although the 90% intervals overlap.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Same basis rules as the category table below
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
6 Claude Haiku 4.5 only
4 Shared
4 Gemini 2.5 Pro only
2 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
SWE-bench VerifiedCoding
Normalized gap 9.5Claude Haiku 4.5:73.3%Gemini 2.5 Pro:63.8%FrontierMath v2 (Tiers 1-3)Math
Normalized gap 8.2Claude Haiku 4.5:5.903%Gemini 2.5 Pro:14.138%FrontierMath v2 (Tier 4)Math
Normalized gap 2.1Claude Haiku 4.5:2.083%Gemini 2.5 Pro:4.167%
Bars run 0–100 on each benchmark’s normalized display scale
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign v5.7 lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
Coding
Like-for-like- Claude Haiku 4.5
- 19.6
- Supported · #124/143
- Gemini 2.5 Pro
- 24.4
- Supported · #108/143
- Basis
- BenchAlign v5.7 lane · 4 vs 3 public rows
- Reading
- Gemini 2.5 Pro leads · intervals overlap
Agentic
Directional only- Claude Haiku 4.5
- 22.0
- Supported · #96/117
- Gemini 2.5 Pro
- 25.4
- Estimated · #91/117
- Basis
- BenchAlign v5.7 lane · 2 vs 1 public rows
- Reading
- Directional only
Knowledge
Directional only- Claude Haiku 4.5
- 34.8
- Estimated · #119/169
- Gemini 2.5 Pro
- 44.1
- Supported · #80/169
- Basis
- BenchAlign v5.7 lane · 2 vs 2 public rows
- Reading
- Directional only
Reasoning
Not comparable- Claude Haiku 4.5
- Not ranked
- Gemini 2.5 Pro
- 69.7
- Unranked · 2 rankable rows
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- Claude Haiku 4.5
- Not ranked
- Gemini 2.5 Pro
- 71.4
- Unranked · 1 rankable row
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- Claude Haiku 4.5
- Not ranked
- Gemini 2.5 Pro
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- Claude Haiku 4.5
- Not ranked
- Gemini 2.5 Pro
- 56.4
- #76/124
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Math
Not comparable- Claude Haiku 4.5
- 28.8
- Unranked · 2 rankable rows
- Gemini 2.5 Pro
- 35.1
- Unranked · 2 rankable rows
- Basis
- Provisional lane · 2 vs 2 weighted rows
- Reading
- Not comparable
| Category | Claude Haiku 4.5 | Gemini 2.5 Pro | Basis | Reading |
|---|---|---|---|---|
| Coding | 19.6Supported · #124/143 | 24.4Supported · #108/143 | Like-for-likeBenchAlign v5.7 lane · 4 vs 3 public rows | Gemini 2.5 Pro leads · intervals overlap |
| Agentic | 22.0Supported · #96/117 | 25.4Estimated · #91/117 | Directional onlyBenchAlign v5.7 lane · 2 vs 1 public rows | Directional only |
| Knowledge | 34.8Estimated · #119/169 | 44.1Supported · #80/169 | Directional onlyBenchAlign v5.7 lane · 2 vs 2 public rows | Directional only |
| Reasoning | Not ranked | 69.7Unranked · 2 rankable rows | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multimodal | Not ranked | 71.4Unranked · 1 rankable row | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Instruction following | Not ranked | 56.4#76/124 | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | 28.8Unranked · 2 rankable rows | 35.1Unranked · 2 rankable rows | Not comparableProvisional lane · 2 vs 2 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.7) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Supported evidence per lane · bars run 0–100Methodology
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- Claude Haiku 4.5
- $0.0035
- Fits in one request
- Gemini 2.5 Pro
- $0.00625
- Fits in one request
Claude Haiku 4.5 has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Claude Haiku 4.5
- $0.065
- Fits in one request
- Gemini 2.5 Pro
- $0.0925
- Fits in one request
Claude Haiku 4.5 has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Claude Haiku 4.5
- $0.09
- Does not fit in one request
- Gemini 2.5 Pro
- $0.15
- Fits in one request
Claude Haiku 4.5 does not fit this workload in one request.
Cached input falls back to the list input rate only where a cached rate is unpublished
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationClaude Haiku 4.5Gemini 2.5 Pro
Context window
Maximum documented context; output-token limits may be lower.
Claude Haiku 4.5
Gemini 2.5 Pro
API model ID
Claude Haiku 4.5
claude-haiku-4-5-20251001
Claude API pricingGemini 2.5 Pro
gemini-2.5-pro
Google Gemini API pricingCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Claude Haiku 4.5
$0.1 per 1M cached input tokens
Claude API pricingGemini 2.5 Pro
$0.125 per 1M cached input tokens
Google Gemini API pricingDocumented inputs
Claude Haiku 4.5
Not sourced
Gemini 2.5 Pro
Not sourced
Documented outputs
Claude Haiku 4.5
Not sourced
Gemini 2.5 Pro
Not sourced
Provider availability
Claude Haiku 4.5
Not sourced
Gemini 2.5 Pro
Not sourced
Reasoning profile
Claude Haiku 4.5
Non-Reasoning
Gemini 2.5 Pro
Non-Reasoning
Weight access
Claude Haiku 4.5
Proprietary
Gemini 2.5 Pro
Proprietary
License
Claude Haiku 4.5
Proprietary
Gemini 2.5 Pro
Proprietary
Release date
Claude Haiku 4.5
2025-10-15
Gemini 2.5 Pro
2025-03-01
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- Gemini 2.5 Pro has the higher public score estimate, 50.17 versus 42.47, but the 90% score intervals overlap.
- Workload cost
- Repository review: $0.065 vs $0.0925. Cache-heavy agent loop: $0.09 vs $0.15.
- Context tradeoff
- Gemini 2.5 Pro has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Questions
Which is better, Claude Haiku 4.5 or Gemini 2.5 Pro?
Gemini 2.5 Pro has the higher public score estimate, 50.17 versus 42.47, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Claude Haiku 4.5 or Gemini 2.5 Pro?
Gemini 2.5 Pro leads the public coding lane, 24.4 to 19.6, with Supported evidence for both models, although the 90% intervals overlap.
Which is better for agentic tasks, Claude Haiku 4.5 or Gemini 2.5 Pro?
Gemini 2.5 Pro scores higher for agentic tasks on the public lane, 25.4 to 22. Gemini 2.5 Pro is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which costs less, Claude Haiku 4.5 or Gemini 2.5 Pro?
For the stated presets, chat costs $0.0035 on Claude Haiku 4.5 and $0.00625 on Gemini 2.5 Pro; repository review costs $0.065 and $0.0925; the cache-heavy agent loop costs $0.09 and $0.15. Claude Haiku 4.5 does not fit this workload in one request.
Which has the larger context window, Claude Haiku 4.5 or Gemini 2.5 Pro?
Gemini 2.5 Pro has the larger documented context window: 1M, compared with 200K.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence14 rows
Agentic
Coding
SWE-bench Verified
Claude Haiku 4.573.3%SourceGemini 2.5 Pro63.8%SourceClaude Haiku 4.5 leads this result
VulcanBench v3
Claude Haiku 4.576.2%SourceGemini 2.5 Pro—Not directly comparable
LiveCodeBench (Vals)
Claude Haiku 4.541.2%SourceGemini 2.5 Pro—Not directly comparable
SWE-bench (Vals)
Claude Haiku 4.566.6%SourceGemini 2.5 Pro54.4%SourceClaude Haiku 4.5 leads this result
Vibe Code Bench
Claude Haiku 4.5—Gemini 2.5 Pro0.40%SourceNot directly comparable
Knowledge
GPQA Diamond (Vals)
Claude Haiku 4.572.2%SourceGemini 2.5 Pro—Not directly comparable
MMLU-Pro (Vals)
Claude Haiku 4.578.7%SourceGemini 2.5 Pro—Not directly comparable
GPQA
Claude Haiku 4.5—Gemini 2.5 Pro83%SourceNot directly comparable
HLE
Claude Haiku 4.5—Gemini 2.5 Pro18.8%SourceNot directly comparable
Math
FrontierMath v2 (Tiers 1-3)
Shared sourceClaude Haiku 4.55.903%Gemini 2.5 Pro14.138%Gemini 2.5 Pro leads this result
FrontierMath v2 (Tier 4)
Shared sourceClaude Haiku 4.52.083%Gemini 2.5 Pro4.167%Gemini 2.5 Pro leads this result
14 public results · 4 shared
Watch Claude Haiku 4.5 vs Gemini 2.5 Pro
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.
Last updated September 29, 2026