Knowledge
Like-for-like- GPT-4.1
- 39.8
- Supported · #133/183
- Grok 4.3
- 57.0
- Supported · #44/183
- Basis
- BenchAlign lane · 2 vs 2 public rows
- Reading
- Grok 4.3 leads · intervals overlap
GPT-4.1 vs Grok 4.3
Updated September 10, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Grok 4.3 has the higher public score estimate, 59.73 versus 42.02, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
1 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Share or export
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.
Chat turn cost
1K fresh input + 500 output tokens
Grok 4.3
Grok 4.3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Grok 4.3
Grok 4.3 has the lower estimated token cost for this stated workload. GPT-4.1 has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback
Show secondary and unsupported calls
Repository review cost
50K fresh input + 3K output tokens
Grok 4.3
Grok 4.3 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
GPT-4.1 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
GPT-4.1 is not ranked on the public lane for agentic, so no winner is named for agentic.
Confidence: limited
Long documents
Prompts that approach the documented context limit
No clear pick
The documented context windows are equal.
Confidence: documented
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
6 GPT-4.1 only
1 Shared
6 Grok 4.3 only
- Shared results
- 1
- GPT-4.1 only
- 6
- Grok 4.3 only
- 6
- Like-for-like categories
- 1 / 8
2 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign 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
Directional only- GPT-4.1
- 41.8
- Estimated · #106/151
- Grok 4.3
- 33.6
- Supported · #131/151
- Basis
- BenchAlign lane · 1 vs 2 public rows
- Reading
- Directional only
Instruction following
Directional only- GPT-4.1
- 50.5
- #82/123
- Grok 4.3
- 94.4
- #2/123
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Directional only
Agentic
Not comparable- GPT-4.1
- Not ranked
- Grok 4.3
- 26.9
- Supported · #144/152
- Basis
- BenchAlign lane · 1 vs 3 public rows
- Reading
- Not comparable
Reasoning
Not comparable- GPT-4.1
- 67.9
- Unranked · 2 rankable rows
- Grok 4.3
- 65.2
- Unranked · 2 rankable rows
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Math
Not comparable- GPT-4.1
- 28.1
- Unranked · 2 rankable rows
- Grok 4.3
- Not ranked
- Basis
- Provisional lane · 2 vs 0 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- GPT-4.1
- Not ranked
- Grok 4.3
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- GPT-4.1
- 50.2
- Unranked · 1 rankable row
- Grok 4.3
- 65.6
- #25/48
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
| Category | GPT-4.1 | Grok 4.3 | Basis | Reading |
|---|---|---|---|---|
| Knowledge | 39.8Supported · #133/183 | 57.0Supported · #44/183 | Like-for-likeBenchAlign lane · 2 vs 2 public rows | Grok 4.3 leads · intervals overlap |
| Coding | 41.8Estimated · #106/151 | 33.6Supported · #131/151 | Directional onlyBenchAlign lane · 1 vs 2 public rows | Directional only |
| Instruction following | 50.5#82/123 | 94.4#2/123 | Directional onlyProvisional lane · 0 vs 0 weighted rows | Directional only |
| Agentic | Not ranked | 26.9Supported · #144/152 | Not comparableBenchAlign lane · 1 vs 3 public rows | Not comparable |
| Reasoning | 67.9Unranked · 2 rankable rows | 65.2Unranked · 2 rankable rows | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | 28.1Unranked · 2 rankable rows | Not ranked | Not comparableProvisional lane · 2 vs 0 weighted rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Multimodal | 50.2Unranked · 1 rankable row | 65.6#25/48 | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
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.
A shared-evidence shape is not available.
BenchLM does not draw a radar or infer missing axes when the matched evidence is too sparse.
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
- GPT-4.1
- $0.006
- Fits in one request
- Grok 4.3
- $0.0025
- Fits in one request
Grok 4.3 has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- GPT-4.1
- $0.124
- Fits in one request
- Grok 4.3
- $0.07
- Fits in one request
Grok 4.3 has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- GPT-4.1
- $0.52
- Fits in one request
- Cached input priced at the published list-input rate
- Grok 4.3
- $0.09
- Fits in one request
Grok 4.3 has the lower modeled cost
GPT-4.1 has no published cached-input rate, so cached tokens use its listed input rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationGPT-4.1Grok 4.3
Context window
Maximum documented context; output-token limits may be lower.
GPT-4.1
1M
Grok 4.3
1M
API model ID
GPT-4.1
Not sourced
Grok 4.3
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
GPT-4.1
Not published
Grok 4.3
$0.2 per 1M cached input tokens
Documented inputs
GPT-4.1
Not sourced
Grok 4.3
Not sourced
Documented outputs
GPT-4.1
Not sourced
Grok 4.3
Not sourced
Provider availability
GPT-4.1
Not sourced
Grok 4.3
Not sourced
Reasoning profile
GPT-4.1
Non-Reasoning
Grok 4.3
Reasoning
Weight access
GPT-4.1
Proprietary
Grok 4.3
Proprietary
License
GPT-4.1
Proprietary
Grok 4.3
Proprietary
Release date
GPT-4.1
2025-04-14
Grok 4.3
2026-04-30
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
- Grok 4.3 has the higher public score estimate, 59.73 versus 42.02, but the 90% score intervals overlap.
- Workload cost
- Repository review: $0.124 vs $0.07. Cache-heavy agent loop: $0.52 vs $0.09.
- Context tradeoff
- Both models list 1M.
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence13 rows
Agentic
Gert Labs
Shared sourceGPT-4.125.65%Grok 4.343.86%Grok 4.3 leads this result
ResearchClawBench
GPT-4.1—Grok 4.312.4%SourceNot directly comparable
Terminal-Bench 2.1 (Vals)
GPT-4.1—Grok 4.341.9%SourceNot directly comparable
Coding
Knowledge
Math
Instruction following
IFEval
GPT-4.187.4%SourceGrok 4.3—Not directly comparable
Frequently asked questions
Which is better, GPT-4.1 or Grok 4.3?
Grok 4.3 has the higher public score estimate, 59.73 versus 42.02, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, GPT-4.1 or Grok 4.3?
GPT-4.1 scores higher for coding on the public lane, 41.8 to 33.6. GPT-4.1 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, GPT-4.1 or Grok 4.3?
GPT-4.1 is not ranked on the public lane for agentic tasks, so no winner is named for agentic tasks.
Which costs less, GPT-4.1 or Grok 4.3?
For the stated presets, chat costs $0.006 on GPT-4.1 and $0.0025 on Grok 4.3; repository review costs $0.124 and $0.07; the cache-heavy agent loop costs $0.52 and $0.09. GPT-4.1 has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, GPT-4.1 or Grok 4.3?
Both models list the same context window, 1M.
Related comparisons
Last updated September 10, 2026
Watch GPT-4.1 vs Grok 4.3
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.