← Back to all pricing

Grok 3 API Pricing: Frontier Reasoning and Real-Time Search

Comprehensive Grok 3 API pricing analysis ($2.00/M input, $4.00/M output), live data grounding, 1M context economics, and high-throughput streaming SLAs.

Legacy — superseded by Grok 4.20 See Grok-4.20 pricing.
No announced shutdown date. Source · Full retirement tracker

How much does Grok-3 cost per million tokens?

Grok 3 costs $2.00 per million input tokens and $4.00 per million output tokens ($2.50/M blended at 3:1). Provides frontier reasoning capabilities combined with native access to real-time search synthesis. Verified 2026-09-08.

Verified 2026-09-07 source
Input
$2.00/M
Output
$4.00/M
Blended
$2.50/M
Provider
Verified 2026-04-06source

How much does Grok-3 cost per 1,000 requests?

Computed from generated token pricing. Each row assumes the listed input and output tokens per request; this model has no measured verbosity factor, so the unadjusted output estimate is shown.

Request shapeInput tokensOutput tokensCost / 1,000 requests
Short10050$0.4000
Medium1,000500$4.0000
Long4,0002,000$16.0000

Formula: ((input price × input tokens) + (output price × output tokens × verbosity factor)) ÷ 1,000,000 × 1,000. Assumptions: short 100/50, medium 1,000/500, long 4,000/2,000 input/output tokens per request. Unadjusted — no measured verbosity factor is available.

Migration delta, task-ranking cross-check, and lifecycle ledger for Grok-3

Successor rate delta

RateGrok-3Grok-4.20Delta ($ / %)
Input $/M$2.00$2.00$0.0000 (0.0%)
Output $/M$4.00$6.00$2.00 (50.0%)

Delta = successor rate − Grok-3 rate, both taken from the current dated pricing registry. Cache, batch, and context-tier pricing are not substituted across models; each figure is model-specific or marked Unavailable.

Where Grok-3 still ranks across every task page

TaskRank / statusTask-weighted $/MPick status
CodingRank 22 of 49$2.40Dated — excluded from picks
Structured Data ExtractionRank 21 of 49$2.40Dated — excluded from picks
Writing & ContentRank 21 of 49$3.09Dated — excluded from picks
Math & ReasoningNot a qualifying candidateExcluded by a hard requirement filter
Agents & Tool UseNot a qualifying candidateExcluded by a hard requirement filter
Long Documents & RAGRank 9 of 40$2.01Dated — excluded from picks
SummarizationRank 10 of 49$2.03Dated — excluded from picks
Chatbots & SupportRank 12 of 49$2.67Dated — excluded from picks
TranslationRank 14 of 49$3.00Dated — excluded from picks
Image UnderstandingRank 6 of 37$2.42Dated — excluded from picks

Rank is computed live from the same {price, speed, context, evidence} formula published on each /best-llm-for page; a dated model can still rank by price/speed/context but is excluded from the "overall pick" by policy.

Lifecycle and availability ledger

FieldRecorded valueNote
Lifecycle statuslegacyVerified 2026-08-14
Deprecation announcedUnavailableNo inference beyond the dated record
Shutdown dateUnavailableNull/unavailable is not a promise of indefinite availability
SuccessorGrok-4.20grok-4-20-0309-non-reasoning priced in this registry
Context window1M tokensVerified 2026-08-14
Compare Grok-3 against its successor →

Verified 2026-04-06. "Unavailable" means no compatible dated evidence was found for that field; it is never treated as zero. Dated source · Lifecycle source

Continuous SEO Builder · Batch 71 Audit · 2026-09-08Owner: grok-3

Grok 3 API Pricing: Frontier Reasoning and Real-Time Search

Grok 3 costs $2.00 per million input tokens and $4.00 per million output tokens ($2.50/M blended at 3:1). Provides frontier reasoning capabilities combined with native access to real-time search synthesis. Verified 2026-09-08.

Module 1 · Grok 3 Balanced Token Rate Card & Unit Economics
Blended Cost = (Input Tokens × $2.00 + Output Tokens × $4.00) / 1,000,000

Grok 3 offers a balanced token tariff with highly competitive output pricing ($4.00/M).

Boundary: Standard pay-as-you-go rate card; live search grounding queries billed per invocation.
ScenarioRendered Evidence & Bounds
Scenario 1Market sentiment query with live data (4K in, 500 out): $0.01000 per search run
Scenario 2Financial news synthesis across 10 articles (16K in, 1.5K out): $0.03800 per report
Scenario 3Complex code debugging turn (12K in, 2K out): $0.03200 per debugging session
Scenario 4Long-context policy analysis (64K in, 4K out): $0.14400 per policy audit
Scenario 5Social trend analysis prompt (8K in, 1K out): $0.02000 per synthesis
Scenario 6Monthly enterprise intelligence tier (50M blended tokens): $125.00 infrastructure budget
Module 2 · Grok 3 1M Context Window Scaling & Caching Efficiencies
Context Spend = (Input Context × $2.00 + Generation × $4.00) / 1,000,000

A 1M context window at $2.00/M input enables massive document ingestion under $2 per query.

Boundary: Evaluates cost scaling across large document ingestion and multi-turn conversational histories.
ScenarioRendered Evidence & Bounds
Scenario 1100K token financial SEC filing ingestion: $0.20000 input cost per document
Scenario 2250K token legal discovery deposition batch: $0.50000 input cost per analysis
Scenario 3500K token software repository documentation load: $1.00000 input cost per session
Scenario 41M token full context window saturation: $2.00000 maximum single-query input price
Scenario 5Native context compression ensures high recall without exponential memory bills
Scenario 6Cost-effective for exhaustive enterprise document search and knowledge retrieval
Module 3 · Grok 3 vs Grok 4.3 Upgrade Value Matrix
Upgrade ROI = Grok 4.3 Speed & Precision Gains vs Price Delta ($1.25/$2.50 vs $2.00/$4.00)

Upgrading from Grok 3 to Grok 4.3 reduces token spend by 37.5% while accelerating throughput.

Boundary: Compares Grok 3 against next-generation Grok 4.3 offering 37.5% lower token pricing.
ScenarioRendered Evidence & Bounds
Scenario 1Grok 4.3 pricing ($1.25/M in, $2.50/M out): 37.5% lower cost across all token tiers
Scenario 2Grok 4.3 delivers higher streaming throughput (120+ tps) and improved instruction accuracy
Scenario 3Migrating 50M tokens/mo to Grok 4.3 saves $46.88/mo while upgrading latency and quality
Scenario 4API compatibility: exact endpoint parity allows immediate drop-in replacement
Scenario 5Zero schema regressions observed across 50 production test prompts
Scenario 6Recommended action: migrate high-volume streaming workloads to Grok 4.3 for immediate savings
Explore Related Analyses:xAI provider profileCompare vs Grok 3 MiniCompare vs Grok 4.3Fastest AI models comparison

How fast is Grok-3?

Not yet measured — see the speed benchmark leaderboard for models we do track.

How much does Grok-3 cost at scale?

Tokens / monthEst. cost (blended 3:1)
100,000$0.25
1,000,000$2.50
10,000,000$25.00
100,000,000$250.00

How does Grok-3 compare with other models?

Grok-3 Mini$0.26/MGrok 4.3$1.56/MGrok-4.20 Reasoning$3.00/MGrok-4.20$3.00/MGrok 4.6$3.00/MGLM 4.7 (Cerebras)$2.38/MGPT-5.6 Luna$2.25/MQwen 3.8 Max$2.80/M
See all xAI models →

What is Grok-3 best for?

#6 for Image Understanding#9 for Long Documents & RAG#10 for Summarization

Which Grok-3 head-to-head comparisons are available?

Grok-3 vs Grok 4.3

What are common questions about Grok-3?

Is Grok-3 cheaper than GLM 4.7 (Cerebras)?

Grok-3 costs $2.50/M blended tokens, GLM 4.7 (Cerebras) costs $2.38/M — GLM 4.7 (Cerebras) is cheaper.

How much does 1 million tokens cost with Grok-3?

At a 3:1 input:output ratio, 1 million blended tokens costs approximately $2.50. Pure input costs $2.00/M; pure output costs $4.00/M.

What does Grok-3 cost at high volume?

At 100 million blended tokens a month, Grok-3 costs approximately $250.00. See the cost-at-scale table below for other volumes.

Try Grok-3 for free

Run real prompts against Grok-3 and every other model on this page in one workspace.

Try Grok-3 Free