Grok 4.5
Opus-class frontier model optimized for coding, agents, and knowledge work at half the cost
Context window
500K
Input / 1M tokens
$2.00
Output / 1M tokens
$6.00
Provider
SpaceXAI (xAI)
Data verified 2026-07-09
Grok 4.5 is SpaceXAI's frontier model trained jointly with Cursor on a 1.5T parameter V9 foundation. Built for coding, agentic tasks, and knowledge work with emphasis on cost-efficiency and token usage. Offers near-frontier capability at significantly lower pricing than competing models like Claude Opus.
Capability index
Relative estimates (0-100) to place this model against its peers, grounded in published benchmarks.
How to access it
Available via SpaceXAI API, Grok Build, Cursor (all plans), X Premium/SuperGrok subscriptions, and third-party integrations. Not yet available in EU.
Strengths
- ✓4.2x more token-efficient than Opus 4.8 on SWE-Bench Pro
- ✓50-60% cheaper than Claude Opus 4.8 and comparable OpenAI models
- ✓Ranks #4 on Artificial Analysis Intelligence Index
- ✓Strong performance on coding and agentic tasks
Best for developers who...
When to choose it (and when not to)
Reach for Grok 4.5 when...
- →When total cost per task is the primary optimization target
- →For multi-step coding and debugging workflows
- →When efficiency per dollar matters more than absolute top-tier benchmark scores
Look elsewhere if...
- ✕When highest raw benchmark performance is required (Fable 5 outperforms on some tasks)
- ✕EU-based teams (not yet available in EU as of launch)
How to use it
- ›Set prompt_cache_key to enable 75% cached input token discount
- ›Route requests consistently to same server for reliable cache hits
- ›Leverage extended reasoning for complex problems (chain-of-thought built-in)
Quickstart
Pythonimport anthropic
client = anthropic.Anthropic(api_key='your-xai-key', base_url='https://api.x.ai')
message = client.messages.create(model='grok-4.5', max_tokens=1024, messages=[{'role': 'user', 'content': 'Write a Python function'}])
print(message.content[0].text)Uses OpenAI-compatible API. Cache keys improve cost efficiency significantly.
API model id: grok-4.5
Benchmarks
| Benchmark | Score | Notes |
|---|---|---|
| GDPval-AA v2 (Agentic Knowledge Work) | 1543 Elo (rank #4) | Behind Fable 5, GPT-5.5, Opus 4.8 |
| Artificial Analysis Intelligence Index | 54 (rank #4) | Well above average; leading models at similar price tier |
| SWE-Bench Pro (Token Efficiency) | 4.2x fewer tokens than Opus 4.8 | Primary efficiency claim; fewer total tokens = lower cost per task |
| Terminal-Bench v2 | Outperforms Opus 4.8 | According to SpaceXAI official benchmarks |
| DeepSWE 1.0 | Ahead of Opus 4.8 | SpaceXAI official data |
| Harvey Legal Agent Benchmark | Top spot | Legal-focused reasoning task |
Source: SpaceXAI official announcement
Compare Grok 4.5
Compare Grok 4.5 with any other model
Build a comparison →All model comparisons →