PREVIOUS FRONTIER GROK500K CONTEXTTEXT + IMAGE INPUTREASONING TO XHIGH2x ABOVE 200K PROMPTUS REGIONS ONLY
Grok 4.6 API Pricing
Grok 4.6 is SpaceXSI's previous frontier model for coding, agentic tasks, and knowledge work, released August 12, 2026 and followed on September 21, 2026 by Grok 4.7, which launched on the same rate card. The pricing page lists $2.00/M input, $0.50/M cached input, and $6.00/M output for prompts under 200K tokens. At or above 200K every rate doubles — $4.00 / $1.00 / $12.00 — and the higher rate applies to the whole request, not just the tokens past the line.
Input - per 1M tokens
$2.00/M
Under 200K prompt live
Output - per 1M tokens
$6.00/M
Reasoning billed as output live
Cached input
$0.50/M
Prompt cache SpaceXSI row -75%
Effective - agentic blend
$1.19/M
92/8 split - 82% cache
§ 01 / TERMINAL
Grok 4.6 API cost calculator.
Live calculator pre-loaded with Grok 4.6's sub-200K rates from SpaceXSI's pricing page. Reasoning effort is configurable across low, medium, high (default) and xhigh, and reasoning tokens bill at the output rate. Note the threshold: send a prompt of 200K tokens or more and every rate in this calculator doubles.
$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
§ 02 / SCENARIOS
Grok 4.6 cost on real workloads.
AGENT
Tool-calling agent
$0.110/task
LONG CONTEXT
Research pack synthesis
$0.199/pack
CHAT
Power-user chat
$0.067/turn
RAG
Knowledge base RAG
$0.188/answer
Estimate · grok-tokenizer-estimate · ≈3.85 chars/token Auto-counts as you type
This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.
Characters 570
Words 93
Tokens (estimated) 148 tokens
Cost as input · uncached $0.0003 USD
Cost as output · uncached $0.00089 USD
Cost as cached input $0.00007 USD
| Model | Input /M | Output /M | Effective blended | Context | Best for |
|---|---|---|---|---|---|
| Grok 4.6 Current | $2.00 cache $0.50 | $6.00 | $1.19 agentic 92/8 | 500K | Coding, agents, and knowledge work |
| Grok 4.7 | $2.00 cache $0.50 | $6.00 | $1.19 same | 500K | Newest Grok, larger base model |
| Grok 4.5 | $2.00 cache $0.30 | $6.00 | $1.04 −13% | 500K | Cache-heavy loops on the prior Grok |
| Grok 4.3 | $1.25 cache $0.20 | $2.50 | $0.558 −2.1× | 1M | Lower-cost Grok chat and tools |
| Grok Build 0.1 | $1.00 cache $0.20 | $2.00 | $0.476 −2.5× | 256K | Dedicated coding workflows |
| Grok 4.20 Multi-Agent (0309) | $1.25 cache $0.20 | $2.50 | $0.558 −2.1× | 1M | Pinned long-horizon agent loops |
| Claude Sonnet 5.5 | $2.00 cache $0.10 | $10.00 | $1.29 +9% | 1M | Claude-class coding agents |
| GPT-5.6 Sol | $4.00 cache $0.40 | $20.00 | $2.73 +2.3× · promo | 1.05M | OpenAI frontier workloads |
Grok 4.6 has published scores on 10 benchmarks we track; its best placing is #28 of 105 on ARC-AGI-2. 4 of these scores come from an independent leaderboard; the rest are the lab's own reports.
| Benchmark | Score | Rank | Leader | Source |
|---|---|---|---|---|
| ARC-AGI-2 Reasoning | 67.1% ARC Prize verified, best result | #28 of 105 | GPT-6 Astra · 95% | Official leaderboard ARC Prize Foundation |
| APEX-Agents Professional work | 57.5% high effort | #2 of 6 | Kimi K3.1 Pro · 69.4% | Vendor-reported SpaceXSI (xAI) |
| Terminal-Bench 3.0 Agents and computer use | 26% high effort | #3 of 9 | DeepSeek V4.1 Flash · 30% | Vendor-reported SpaceXSI (xAI) |
| ARC-AGI-3 Reasoning · Standard harness | 2.1% ARC Prize verified, best result | #9 of 24 | GPT-6 Astra · 62.7% | Official leaderboard ARC Prize Foundation |
| ARC-AGI-1 Reasoning | 87.5% ARC Prize verified, best result | #38 of 99 | 6 tied · 98.5% | Official leaderboard ARC Prize Foundation |
| FrontierCode Coding · v1.1 Extended (all 150 tasks) | 61.3% high effort | #2 of 4 | GPT-6 Astra · 64.5% | Vendor-reported SpaceXSI (xAI) |
| Harvey Legal Agent Benchmark Professional work | 15.8% high effort, run by Vals | #4 of 7 | Kimi K3.1 Pro · 23.7% | Vendor-reported SpaceXSI (xAI) |
| DeepSWE Coding | 65.9% high effort | #16 of 27 | Gemini 4 Argon · 77.9% | Vendor-reported SpaceXSI (xAI) |
| Terminal-Bench 4.0 Agents and computer use | 20.3% high effort, Grok Build harness, 330 trials, ±3.1 pts (95% CI) | #24 of 34 | Kimi K3.1 Pro · 65.7% | Official leaderboard Terminal-Bench (Stanford, Laude Institute) |
| CursorBench 4.0 Coding | 40.4% high effort | #6 of 7 | Claude Opus 5.5 · 57.8% | Vendor-reported SpaceXSI (xAI) |
Grok 4.6 API pricing FAQ.
Grok 4.6 pricing questions, with SpaceXSI's token rates separated from the long-context threshold and cache assumptions.
Q · 01 What is Grok 4.6's API price? +
SpaceXSI lists
grok-4.6 at $2.00/M input, $0.50/M cached input, and $6.00/M output for prompts below 200K tokens. These are USD prices per 1M tokens.Q · 02 Why does Grok 4.6 have two sets of prices? +
SpaceXSI charges long-context requests at double the rate:
$4.00/M input, $1.00/M cached input, and $12.00/M output. The trigger is the prompt reaching 200K tokens, and the higher rate then applies to every token in that request — not only the tokens beyond 200K. A 199K-token prompt and a 201K-token prompt therefore differ in cost by roughly 2x, not by 1%.Q · 03 Is Grok 4.6 more expensive than Grok 4.5? +
On paper they are identical: both list
$2/M input and $6/M output. The difference is cache. Grok 4.5's cached input was cut to $0.30/M on the day Grok 4.6 launched, while Grok 4.6 reads cache at $0.50/M. Under our standard blend (92% input, 82% cache hits) that puts Grok 4.6 at $1.19/M effective against Grok 4.5's $1.04/M — so on cache-heavy agent loops the older model is currently the cheaper one.Q · 04 What context window does Grok 4.6 have? +
The SpaceXSI model page lists
500,000 tokens, the same window as Grok 4.5. Grok 4.3 and the Grok 4.20 variants are listed at 1M, so Grok 4.6 is not the longest-context Grok. Grok 4.7 keeps the same 500K window.Q · 05 Does Grok 4.6 support batch pricing? +
Yes, 20%, since late October 2026. Between October 19 and 21 SpaceXSI added
grok-4.6 to its batch list without a release note, and the model page, which said Batch API: Not supported in September, now reads Supported: $1.60/M input and $4.80/M output, 20% off every token type. Priority processing bills at 2x standard rates.Q · 06 Where can I call Grok 4.6? +
The model page lists regions
us-east-1, us-west-2 and us-central-1, with rate limits of 150 requests per second and 50M tokens per minute. It supports function calling, structured outputs, and reasoning at low, medium, high (default), and xhigh effort. Grok 4.6 is also served on SpaceXSI's US regional endpoint, us.api.x.ai, which keeps inference in the United States and bills every rate at 1.1x.Q · 07 Should I move from Grok 4.6 to Grok 4.7? +
Price does not decide it. Grok 4.7 launched on this model's card, and SpaceXSI describes it as a new, larger base model that is better at checking its own work and at long context. Grok 4.6 stays on the pricing page, and SpaceXSI has published no retirement date for it, so the case for staying is pinning behaviour you have already evaluated.