Skip to main content

    xAI

    Grok 4.6

    xAI's August 12, 2026 model for coding, agentic tasks and knowledge work. It builds on Grok 4.5 with supplemental training, regenerated supervised fine-tuning trajectories and agentic reinforcement learning. At launch, xAI reported DeepSWE v1.1 65.9% and APEX-Agents 57.5% at high reasoning effort. It accepts text and images, produces text and has a 500K-token context window, with low/medium/high/xhigh reasoning effort. Standard xAI API input/output rates are $2/$6 per MTok (input/output) below 200K prompt tokens and $4/$12 at or above 200K; cached input is $0.50/$1 respectively. Microsoft Foundry availability was announced August 26, 2026.

    500K tokens
    Multimodal
    Max
    New
    Provider
    xAI
    Context
    500K
    Catalog · As of Aug 13, 2026Within 30d
    yno subscription tier
    Max
    Released
    Aug 12, 2026
    Speed
    Medium
    Reasoning
    Expert
    Modality
    Multimodal

    Benchmarks

    Quality

    AA Intelligence Index
    44.4points
    Artificial Analysis · Date unknownAge unknown
    AA Coding Index
    76.8points
    Artificial Analysis · Date unknownAge unknown
    GPQA Diamond
    94.9%
    Artificial Analysis · Date unknownAge unknown

    Speed and latency

    Output speed
    67.86tokens/s
    Configuration: Prompt length: 1,000 · Parallel queries: 1
    Artificial Analysis · Date unknownAge unknown
    Time to first token
    37.39s
    Configuration: Prompt length: 1,000 · Parallel queries: 1
    Artificial Analysis · Date unknownAge unknown
    Time to first answer token
    37.39s
    Configuration: Prompt length: 1,000 · Parallel queries: 1
    Artificial Analysis · Date unknownAge unknown

    API pricing

    API input cost
    $2.00/ 1M tokens
    Artificial Analysis · Date unknownAge unknown
    API output cost
    $6.00/ 1M tokens
    Artificial Analysis · Date unknownAge unknown

    Benchmark variant: Grok 4.6 (high)

    AA data retrieved Sep 11, 2026 · Artificial Analysis

    Indexes use points; evaluations use accuracy percentages. API measurements do not measure yno application performance. Methodology

    Capabilities

    • Reasoning
    • Code
    • Agentic
    • Vision
    • Long Horizon
    • Tool Use

    Use cases

    • Long-running coding agents
    • Agentic knowledge work
    • Terminal automation
    • Tool-calling agents

    Strengths

    • Long-running agents
    • xAI-reported DeepSWE v1.1 65.9% at launch
    • Self-testing and verification
    • Four reasoning-effort levels

    Best for

    Teams that want frontier long-horizon agent capability at xAI's aggressive pricing

    Use Grok 4.6 in yno.ai

    No credit card required

    Related models

    Frequently asked

    What is Grok 4.6?
    xAI's August 12, 2026 model for coding, agentic tasks and knowledge work. It builds on Grok 4.5 with supplemental training, regenerated supervised fine-tuning trajectories and agentic reinforcement learning. At launch, xAI reported DeepSWE v1.1 65.9% and APEX-Agents 57.5% at high reasoning effort. It accepts text and images, produces text and has a 500K-token context window, with low/medium/high/xhigh reasoning effort. Standard xAI API input/output rates are $2/$6 per MTok (input/output) below 200K prompt tokens and $4/$12 at or above 200K; cached input is $0.50/$1 respectively. Microsoft Foundry availability was announced August 26, 2026.
    How much does Grok 4.6 cost in yno.ai?
    Grok 4.6 is available on the Max tier of yno.ai.
    What can Grok 4.6 do?
    Grok 4.6 is best for Teams that want frontier long-horizon agent capability at xAI's aggressive pricing. Its main capabilities include Reasoning, Code, Agentic, Vision, Long Horizon, Tool Use.
    How does Grok 4.6 compare to other models?
    Grok 4.6 excels at Long-running agents and is recommended for Long-running coding agents, Agentic knowledge work, Terminal automation, Tool-calling agents. See related models below.
    How do I use Grok 4.6 in yno.ai?
    Sign up for yno.ai, select Grok 4.6 from the model picker, and start chatting.