vs
0:2
overall capability wins (0 ties across 2 scored). Grok 4.5 leads overall.
- Pricing3:0
- Speed2:0
- Specs0:2
Verdict
Grok 4.5 offers ~3.8x larger context (500K vs 131K); Grok 4.5 allows ~15.6x more max output (128K vs 8K); Grok 3 Mini is ~8.6x cheaper on a blended token basis than Grok 4.5.
- Grok 4.5 offers ~3.8x larger context (500K vs 131K)
- Grok 4.5 allows ~15.6x more max output (128K vs 8K)
- Grok 3 Mini is ~8.6x cheaper on a blended token basis than Grok 4.5
Cost for 1M in + 250K out
Illustrative chat workload at primary-provider list prices.
- Grok 3 Mini
- $0.42
- Grok 4.5
- $3.50
Grok 3 Mini is about 8.2x cheaper on this workload.
Price per 1M tokens
Full comparison
Organization
xAI
xAI
Family
Grok
Grok
License
Proprietary
Proprietary
Open weights
No
No
Release
Feb 19, 2025
Jul 9, 2026
Knowledge cutoff
-
-
API / provider
xAI
xAI
Modalities
text → text
text, image → text
Specs
Context window
131K
500K
Max output
8K
128K
Parameters
—
—
Pricing
Input $/1M
$0.30
$2
Output $/1M
$0.50
$6
Blended $/1M (3∶1)
$0.35
$3
Speed
tok/s
130
100
TTFT (s)
0.22
0.35
Reasoning
MMLU-Pro
—
—
GPQA Diamond
—
93.1
Humanity's Last Exam
—
42.7
AIME 2025
—
—
MATH-500
—
—
Humanity's Last Exam (with tools)
—
—
AA-Omniscience Accuracy
—
51.6
AA-LCR v1.1
—
79.3
CritPt
—
15.4
MMMU-Pro
—
80.4
IFBench
—
—
Chartography
—
—
Chartography (With Tools)
—
—
Coding
SWE-bench Verified
—
—
SWE-bench Pro
—
—
SWE-bench Multilingual
—
—
LiveCodeBench
—
—
Terminal-Bench 2.1
—
81.6
Aider Polyglot
—
—
Terminal-Bench 3
—
15.7
BigCodeBench
—
—
SciCode
—
55.0
CursorBench
—
—
SWE-Rebench
—
63.8
NL2Repo-Bench
—
—
DeepSWE
—
—
WebDev Arena
—
1,555
Terminal-Bench 4.0
—
12.4
CursorBench 4.0
—
—
FrontierCode v1.1 (Main)
—
—
Arena
LMArena Elo
—
1,468
| Metric | Grok 3 Mini | Grok 4.5 | Delta |
|---|---|---|---|
| Identity | |||
| Organization | xAI | xAI | — |
| Family | Grok | Grok | — |
| License | Proprietary | Proprietary | — |
| Open weights | No | No | — |
| Release | Feb 19, 2025 | Jul 9, 2026 | — |
| Knowledge cutoff | - | - | — |
| API / provider | xAI | xAI | — |
| Modalities | text → text | text, image → text | — |
| Specs | |||
| Context window | 131K | 500K | B +369K |
| Max output | 8K | 128K | B +120K |
| Parameters | — | — | — |
| Pricing | |||
| Input $/1M | $0.30 | $2 | A +$1.70 |
| Output $/1M | $0.50 | $6 | A +$5.50 |
| Blended $/1M (3∶1) | $0.35 | $3 | A +$2.65 |
| Speed | |||
| tok/s | 130 | 100 | A +30 |
| TTFT (s) | 0.22 | 0.35 | A +0.13 |
| Reasoning | |||
| MMLU-Pro | — | — | — |
| GPQA Diamond | — | 93.1 | — |
| Humanity's Last Exam | — | 42.7 | — |
| AIME 2025 | — | — | — |
| MATH-500 | — | — | — |
| Humanity's Last Exam (with tools) | — | — | — |
| AA-Omniscience Accuracy | — | 51.6 | — |
| AA-LCR v1.1 | — | 79.3 | — |
| CritPt | — | 15.4 | — |
| MMMU-Pro | — | 80.4 | — |
| IFBench | — | — | — |
| Chartography | — | — | — |
| Chartography (With Tools) | — | — | — |
| Coding | |||
| SWE-bench Verified | — | — | — |
| SWE-bench Pro | — | — | — |
| SWE-bench Multilingual | — | — | — |
| LiveCodeBench | — | — | — |
| Terminal-Bench 2.1 | — | 81.6 | — |
| Aider Polyglot | — | — | — |
| Terminal-Bench 3 | — | 15.7 | — |
| BigCodeBench | — | — | — |
| SciCode | — | 55.0 | — |
| CursorBench | — | — | — |
| SWE-Rebench | — | 63.8 | — |
| NL2Repo-Bench | — | — | — |
| DeepSWE | — | — | — |
| WebDev Arena | — | 1,555 | — |
| Terminal-Bench 4.0 | — | 12.4 | — |
| CursorBench 4.0 | — | — | — |
| FrontierCode v1.1 (Main) | — | — | — |
| Arena | |||
| LMArena Elo | — | 1,468 | — |
Grok 3 Mini
- Context
- 131K / 8K out
- Parameters
- —
- Price
- $0.30 / $0.50
- Speed
- 130 tok/s · 0.22s TTFT
- Modalities
- text → text
- License
- Proprietary
Reasoning-capable mini tier from the Grok 3 generation — strong STEM relative to cost.
Grok 4.5
- Context
- 500K / 128K out
- Parameters
- —
- Price
- $2 / $6
- Speed
- 100 tok/s · 0.35s TTFT
- Modalities
- text, image → text
- License
- Proprietary
xAI/SpaceXAI public API flagship (July 2026) - Cursor-trained coding focus at $2/$6. For the Cursor IDE jointly-trained SKU and Fast tier, see Cursor Grok 4.5.
Benchmark charts
Winner bars are emphasized. Per-benchmark deltas sit above each chart.
These two models have no overlapping published benchmarks in our dataset. Compare specs and pricing instead.