SpaceXAI: Grok 4.3 vs SpaceXAI: Grok 4.20
Side-by-side comparison of specs, pricing, benchmark scores, and task rankings. Updated 2026-11-07.
| SpaceXAI: Grok 4.3 | SpaceXAI: Grok 4.20 | |
|---|---|---|
| Vendor | x-ai | x-ai |
| Quality Score | 100 | 100 |
| Benchmark Score | 58.6 | 56.5 |
| Input Price | $1.25/M | $1.25/M |
| Output Price | $2.50/M | $2.50/M |
| Context Window | 1,000,000 | 2,000,000 |
| Max Output | 900,000 | 1,800,000 |
| Tool Calling | ✓ | ✓ |
| Structured Output | ✓ | ✓ |
| Reasoning Mode | ✓ | ✓ |
| Vision | ✓ | ✓ |
| Audio | - | - |
| Benchmark Scores | ||
| ai_index | 62.0 | 61.0 |
| ai_index_agentic | 39.8 | - |
| ai_index_coding | 69.7 | - |
| eqbench | - | 45.7 |
The verdict
For a sample job of 1 million input tokens and 250,000 output tokens, both cost $1.88 at current list prices.
Grok 4.20 reads up to 2,000,000 tokens at once, roughly 3,000 pages of text, against about 1,500 pages for Grok 4.3.
On our blended benchmark score Grok 4.3 leads 58.6 to 56.5.
Across the 20 tasks where they differ, Grok 4.3 scores higher on 17 (SQL Generation, Code Review, Code Refactoring and others), Grok 4.20 on 3 (Code Completion, Code Documentation, Long-Document Summarization).
Grok 4.3 was released in April 2026 and Grok 4.20 in March 2026.
Pick Grok 4.3 if you need benchmark performance or the stronger all-round task scores; pick Grok 4.20 if you need long documents.
Costs use current list prices per million tokens; page counts assume about 0.75 words per token and 500 words per page.
Who wins by task?
| Task | SpaceXAI: Grok 4.3 | SpaceXAI: Grok 4.20 |
|---|---|---|
| SQL Generation | 158 | 144 |
| Code Review | 155 | 149 |
| Code Completion | 120 | 121 |
| Code Refactoring | 155 | 151 |
| Bug Fixing | 164 | 153 |
| Unit Test Generation | 144 | 134 |
| Code Documentation | 139 | 140 |
| Regex Writing | 130 | 127 |
| CI/CD Pipelines | 136 | 130 |
| Frontend Component Design | 138 | 131 |
| Data Analysis | 153 | 135 |
| CSV / Spreadsheet Cleanup | 147 | 138 |
| ETL Scripting | 145 | 141 |
| JSON Extraction | 134 | 123 |
| Bulk Data Labeling | 124 | 120 |
| OCR / Document Parsing | 141 | 135 |
| Table Extraction from PDFs | 141 | 135 |
| Long-Document Summarization | 152 | 153 |
| Short-Form Summarization | 120 | 119 |
| Blog Post Writing | 134 | 132 |
Scores reflect capability match + benchmark data + pricing for each task. Methodology →