Skip to content

Category

Language

General text generation and understanding. The workhorse category: drafting, summarising, extraction and conversation across every domain. Start here for anything text-shaped and trade up only when a task genuinely needs it. The deciding numbers are usually price per million and how much of the you actually fill.

17 models in Language from 1 provider

Top models

Leading language models

View all →
  • AC Flash 4.5 by Webparam
    Webparam

    AC Flash 4.5

    webparam/ac-flash-4-5-standard

    Fast tier — the cheapest way to run high volume. Measured at 4.6s to first token and 14.5 tokens/sec.

    LanguageTool callingJSON modeStreaming

    — ctx · $10 in · $40 out

  • AC Pro 4.6 by Webparam
    Webparam

    AC Pro 4.6

    webparam/ac-pro-4-6-standard

    Pro tier — stronger reasoning where quality matters more than cost. Measured at 3.7s to first token and 20.7 tokens/sec.

    LanguageTool callingJSON modeStreaming

    — ctx · $60 in · $250 out

  • AC Pro 5 by Webparam
    Webparam

    AC Pro 5

    webparam/ac-pro-5-standard

    Pro tier — stronger reasoning where quality matters more than cost. Measured at 4.0s to first token and 30.3 tokens/sec.

    LanguageTool callingJSON modeStreaming

    — ctx · $60 in · $250 out

  • AC Supreme Safe 5 by Webparam
    Webparam

    AC Supreme Safe 5

    webparam/ac-supreme-safe-5-standard

    Ultra tier — the most capable models on the platform. Measured at 4.0s to first token and 30.4 tokens/sec.

    LanguageTool callingJSON modeStreaming

    — ctx · $100 in · $400 out

Browse all language models →

17 models in the directory

How to choose

Choosing a model in Language

  1. 01

    Match cost to volume

    High-volume pipelines rarely need flagship pricing. Mid-tier models handle most drafting and summarisation at a fraction of the cost.

  2. 02

    Check context honestly

    A huge context window matters only if you actually fill it. Paying for 1M tokens of context you never use is the most common overspend.

  3. 03

    Test your own prompts

    Leaderboards measure averages. Your production prompts are not average. Try two or three candidates side by side before committing.

Explore nearby

Where to look next

All categories →