none reviewed yet · 6 benchmark observations · latest capture Sep 8, 2026
Inspect the evidencexAI · MODEL PROFILE
Grok 4.5
Teams that need a documented SWE-bench Pro number from xAI's line rather than 4.6's unpublished claims
Saved editorial fit for your current task lens; not a benchmark result
How scoring worksVerified rates (Sep 8, 2026) · 100 runs / month at your token settings
Adjust workloadAt a glance
9 facts- Provider
- xAI
- Type
- API model
- Access
- xAI API and console, Grok Build (was the default model), Cursor on all plans
- Context
- 500,000 tokens
- Maximum output
- n.a. on fetched pages
- License / weights
- Proprietary
- Input modalities
- text, image, file
- Output modalities
- text
- Pricing notes
- $2 in / $6 out per 1M; cached input $0.30 (BenchLM notes this is a 10% cache-read estimate, not an xAI-published rate); higher-context pricing applies above 200K with an undisclosed surcharge
The practical trade-offs
Documented strengths
- SWE-bench Pro 64.7% (rank 10) beats GPT-5.6 Sol's 64.6% and Opus 4.7's 64.3%
- Roughly 2x token efficiency and under half the steps of comparable models per xAI
- Well-sampled Arena row (6,034 votes) and Design Arena 1301
- Same $2/$6 price as Grok 4.6
Limitations to check
- Superseded by Grok 4.6, which scores higher on both preference boards at the same price
- AA publishes two conflicting Intelligence Index values (45 vs 56)
- 500K context with an undisclosed >200K surcharge
- Cached-input rate is an estimate, not an xAI-published price
Follow the source
6 source linksEvidence & conditions
6Source links
NoneReviewed records
6Benchmark runs
Sep 8, 2026Latest capture
- Evidence source 1
x.aiDate not recorded - Evidence source 2
benchlm.aiDate not recorded - Evidence source 3
arena.aiDate not recorded - Evidence source 4
benchlm.aiDate not recorded - Evidence source 5
benchlm.aiDate not recorded - Evidence source 6
artificialanalysis.aiDate not recorded
Benchmark observations
SWE-bench Pro Inspect conditions
- name
- SWE-bench Pro
- value
- 64.7
- unit
- %
- source Url
- https://x.ai/news/grok-4-5
- checked At
- 2026-09-08T19:31:52Z
- attribution
- provider-reported
- configuration
- Resolve rate; Grok 4.5 launch evaluation; scaffold and effort unspecified.
Terminal-Bench 2.1 Inspect conditions
- name
- Terminal-Bench 2.1
- value
- 83.3
- unit
- %
- source Url
- https://x.ai/news/grok-4-5
- checked At
- 2026-09-08T19:31:52Z
- attribution
- provider-reported
- configuration
- Grok 4.5 launch evaluation; scaffold and effort unspecified.
DeepSWE 1.0 Inspect conditions
- name
- DeepSWE 1.0
- value
- 62
- unit
- %
- source Url
- https://x.ai/news/grok-4-5
- checked At
- 2026-09-08T19:31:52Z
- attribution
- provider-reported
- configuration
- Datacurve evaluation, run by Artificial Analysis using provider-specific harnesses.
DeepSWE 1.1 Inspect conditions
- name
- DeepSWE 1.1
- value
- 53
- unit
- %
- source Url
- https://x.ai/news/grok-4-5
- checked At
- 2026-09-08T19:31:52Z
- attribution
- provider-reported
- configuration
- Datacurve run with mini-swe-agent; Grok 4.5 launch result, not the later High row.
SWE Marathon Inspect conditions
- name
- SWE Marathon
- value
- 29
- unit
- %
- source Url
- https://x.ai/news/grok-4-5
- checked At
- 2026-09-08T19:31:52Z
- attribution
- provider-reported
- configuration
- Resolution rate, pass@1; Grok 4.5 launch evaluation.
Artificial Analysis Intelligence Index Inspect conditions
- name
- Artificial Analysis Intelligence Index
- value
- 56
- unit
- index
- source Url
- https://x.ai/news/grok-4-6
- checked At
- 2026-09-08T19:31:52Z
- attribution
- provider-reported
- configuration
- Grok 4.5 High comparison row; reported by xAI, not a fresh independent AA retrieval.
Source reports, measured results, and editorial assessments are separate evidence types. Source dates do not imply a new benchmark run. Capture dates are shown when recorded; otherwise we say the capture date was not recorded. Recorded dates are compared with this catalogue snapshot (Sep 11, 2026), not today’s date.
Method AIS-V2
How this page is assessed
Quality claims follow Method AIS-V2.
Money never buys rank — commercial relationships cannot change rank, eligibility, evidence quality, or publication timing.
Submit or claim a listing is free and does not skip the editorial queue.
Private · this device only