Z.ai: GLM 4.6 (exacto)

Description

Compared with GLM-4.5, this generation brings several key improvements: Longer context window: The context window has been expanded from 128K to 200K tokens, enabling the model to handle more complex agentic tasks. Superior coding performance: The model achieves higher scores on code benchmarks and demonstrates better real-world performance in applications such as Claude Code、Cline、Roo Code and Kilo Code, including improvements in generating visually polished front-end pages. Advanced reasoning: GLM-4.6 shows a clear improvement in reasoning performance and supports tool use during inference, leading to stronger overall capability. More capable agents: GLM-4.6 exhibits stronger performance in tool using and search-based agents, and integrates more effectively within agent frameworks. Refined writing: Better aligns with human preferences in style and readability, and performs more naturally in role-playing scenarios.

Raw fields snapshot

{
  "id": "z-ai/glm-4.6:exacto",
  "name": "Z.ai: GLM 4.6 (exacto)",
  "description": "Compared with GLM-4.5, this generation brings several key improvements:\n\nLonger context window: The context window has been expanded from 128K to 200K tokens, enabling the model to handle more complex agentic tasks.\nSuperior coding performance: The model achieves higher scores on code benchmarks and demonstrates better real-world performance in applications such as Claude Code、Cline、Roo Code and Kilo Code, including improvements in generating visually polished front-end pages.\nAdvanced reasoning: GLM-4.6 shows a clear improvement in reasoning performance and supports tool use during inference, leading to stronger overall capability.\nMore capable agents: GLM-4.6 exhibits stronger performance in tool using and search-based agents, and integrates more effectively within agent frameworks.\nRefined writing: Better aligns with human preferences in style and readability, and performs more naturally in role-playing scenarios.",
  "created": 1759235576,
  "canonical_slug": "z-ai/glm-4.6",
  "hugging_face_id": "",
  "source_type": "both",
  "context_length": 204800,
  "max_completion_tokens": 131072,
  "is_moderated": false,
  "architecture": {
    "modality": "text->text",
    "input_modalities": [
      "text"
    ],
    "output_modalities": [
      "text"
    ],
    "tokenizer": "Other",
    "instruct_type": null
  },
  "input_modalities": [
    "text"
  ],
  "output_modalities": [
    "text"
  ],
  "modality": "text->text",
  "tokenizer": "Other",
  "instruct_type": null,
  "supported_parameters": [
    "frequency_penalty",
    "include_reasoning",
    "max_tokens",
    "presence_penalty",
    "reasoning",
    "repetition_penalty",
    "response_format",
    "seed",
    "stop",
    "structured_outputs",
    "temperature",
    "tool_choice",
    "tools",
    "top_k",
    "top_p"
  ],
  "default_parameters": {
    "temperature": 0.6,
    "top_p": null,
    "frequency_penalty": null
  },
  "per_request_limits": null,
  "top_provider": {
    "context_length": 204800,
    "max_completion_tokens": 131072,
    "is_moderated": false
  },
  "pricing": {
    "prompt": "0.00000044",
    "completion": "0.00000176",
    "input_cache_read": "0.00000011"
  },
  "PPM": {
    "prompt": 0.44,
    "completion": 1.76,
    "input_cache_read": 0.11
  },
  "openrouter_raw": {
    "id": "z-ai/glm-4.6:exacto",
    "canonical_slug": "z-ai/glm-4.6",
    "hugging_face_id": "",
    "name": "Z.ai: GLM 4.6 (exacto)",
    "created": 1759235576,
    "description": "Compared with GLM-4.5, this generation brings several key improvements:\n\nLonger context window: The context window has been expanded from 128K to 200K tokens, enabling the model to handle more complex agentic tasks.\nSuperior coding performance: The model achieves higher scores on code benchmarks and demonstrates better real-world performance in applications such as Claude Code、Cline、Roo Code and Kilo Code, including improvements in generating visually polished front-end pages.\nAdvanced reasoning: GLM-4.6 shows a clear improvement in reasoning performance and supports tool use during inference, leading to stronger overall capability.\nMore capable agents: GLM-4.6 exhibits stronger performance in tool using and search-based agents, and integrates more effectively within agent frameworks.\nRefined writing: Better aligns with human preferences in style and readability, and performs more naturally in role-playing scenarios.",
    "context_length": 204800,
    "architecture": {
      "modality": "text->text",
      "input_modalities": [
        "text"
      ],
      "output_modalities": [
        "text"
      ],
      "tokenizer": "Other",
      "instruct_type": null
    },
    "pricing": {
      "prompt": "0.00000044",
      "completion": "0.00000176",
      "input_cache_read": "0.00000011"
    },
    "top_provider": {
      "context_length": 204800,
      "max_completion_tokens": 131072,
      "is_moderated": false
    },
    "per_request_limits": null,
    "supported_parameters": [
      "frequency_penalty",
      "include_reasoning",
      "max_tokens",
      "presence_penalty",
      "reasoning",
      "repetition_penalty",
      "response_format",
      "seed",
      "stop",
      "structured_outputs",
      "temperature",
      "tool_choice",
      "tools",
      "top_k",
      "top_p"
    ],
    "default_parameters": {
      "temperature": 0.6,
      "top_p": null,
      "frequency_penalty": null
    },
    "expiration_date": null
  },
  "categories": {
    "text-expert": {
      "score": 1435,
      "rank_ub": 49,
      "votes": 1834,
      "ci_95": "±14"
    },
    "text-industry-software-and-it-services": {
      "score": 1455,
      "rank_ub": 47,
      "votes": 12517,
      "ci_95": "±6"
    },
    "text-industry-writing-and-literature-and-language": {
      "score": 1403,
      "rank_ub": 45,
      "votes": 7604,
      "ci_95": "±7"
    },
    "text-overall": {
      "score": 1425,
      "rank_ub": 39,
      "votes": 35128,
      "ci_95": "±4"
    },
    "text-industry-life-and-physical-and-social-science": {
      "score": 1448,
      "rank_ub": 39,
      "votes": 5742,
      "ci_95": "±8"
    },
    "text-industry-mathematical": {
      "score": 1444,
      "rank_ub": 27,
      "votes": 1678,
      "ci_95": "±14"
    },
    "text-industry-entertainment-and-sports-and-media": {
      "score": 1396,
      "rank_ub": 43,
      "votes": 6251,
      "ci_95": "±8"
    },
    "text-industry-business-and-management-and-financial-operations": {
      "score": 1409,
      "rank_ub": 56,
      "votes": 6812,
      "ci_95": "±8"
    },
    "text-industry-medicine-and-healthcare": {
      "score": 1448,
      "rank_ub": 49,
      "votes": 1827,
      "ci_95": "±14"
    },
    "text-industry-legal-and-government": {
      "score": 1435,
      "rank_ub": 42,
      "votes": 2324,
      "ci_95": "±13"
    },
    "text-instruction-following": {
      "score": 1415,
      "rank_ub": 41,
      "votes": 9843,
      "ci_95": "±6"
    },
    "text-math": {
      "score": 1424,
      "rank_ub": 40,
      "votes": 2090,
      "ci_95": "±13"
    },
    "text-creative-writing": {
      "score": 1401,
      "rank_ub": 38,
      "votes": 4706,
      "ci_95": "±9"
    },
    "text-multi-turn": {
      "score": 1421,
      "rank_ub": 49,
      "votes": 5683,
      "ci_95": "±8"
    },
    "text-hard-prompts": {
      "score": 1442,
      "rank_ub": 45,
      "votes": 18861,
      "ci_95": "±5"
    },
    "text-coding": {
      "score": 1460,
      "rank_ub": 53,
      "votes": 7479,
      "ci_95": "±7"
    },
    "text-hard-prompts-english": {
      "score": 1450,
      "rank_ub": 45,
      "votes": 9365,
      "ci_95": "±7"
    },
    "text-longer-query": {
      "score": 1433,
      "rank_ub": 43,
      "votes": 8399,
      "ci_95": "±7"
    },
    "text-english": {
      "score": 1439,
      "rank_ub": 40,
      "votes": 16732,
      "ci_95": "±5"
    },
    "text-chinese": {
      "score": 1482,
      "rank_ub": 19,
      "votes": 1670,
      "ci_95": "±15"
    },
    "text-french": {
      "score": 1449,
      "rank_ub": 32,
      "votes": 635,
      "ci_95": "±26"
    },
    "text-spanish": {
      "score": 1412,
      "rank_ub": 42,
      "votes": 888,
      "ci_95": "±22"
    },
    "text-russian": {
      "score": 1407,
      "rank_ub": 49,
      "votes": 2773,
      "ci_95": "±11"
    },
    "text-german": {
      "score": 1423,
      "rank_ub": 28,
      "votes": 631,
      "ci_95": "±24"
    },
    "text-japanese": {
      "score": 1373,
      "rank_ub": 35,
      "votes": 299,
      "ci_95": "±35"
    },
    "text-korean": {
      "score": 1367,
      "rank_ub": 36,
      "votes": 530,
      "ci_95": "±27"
    },
    "text-exclude-ties": {
      "score": 1416,
      "rank_ub": 41,
      "votes": 24568,
      "ci_95": "±6"
    },
    "code-overall": {
      "score": 1355,
      "rank_ub": 25,
      "votes": 8747,
      "ci_95": "+8/-8"
    },
    "code-html": {
      "score": 1362,
      "rank_ub": 25,
      "votes": 8715,
      "ci_95": "+9/-9"
    }
  },
  "arena_model_id": "z-ai/glm-4.6",
  "leaderboard_name": "glm-4.6",
  "match_method": "openrouter_id",
  "match_key": "z-ai/glm-4.6",
  "match_input": "z-ai/glm-4.6:exacto",
  "match_used_alias": "z-ai/glm-4.6:exacto",
  "match_resolved_to": "z-ai/glm-4.6",
  "arena_aliases": []
}