{ "resource_id": "deepseek-v4-pro", "type": { "resource_type": "Model", "resource_tag": [ "1M Context", "1.32", "3.96", ".044" ] }, "capability": { "summary": "Text reasoning model from the larger sparse-expert architecture tier.", "core_primitives": [ "mathematical reasoning", "code generation", "document analysis", "tool-assisted problem solving" ], "problem_space": "Text model for analytical and programming tasks. Applicable tasks include mathematical reasoning, code generation, document analysis, tool-assisted problem solving. Accepts text input and produces text output. External tools are required to execute programs, browse sources, or modify files; generated instructions alone do not perform those actions. Published distinction: The published family architecture has 1.6 trillion total and 49 billion active parameters. The configured dated supplier route corresponds by name to the August agent-focused update; greater model size is not a measured quality guarantee.", "domain_tags": [ "Language", "Software Engineering" ] }, "constraint": { "env_requirements": [ "DeepSeek-Compatible API Endpoint", "Valid API Key (Cloud)", "Stable Network Connection" ], "io_signature": "Input modalities: text | Output modalities: text. Gateway transport support requires verification." }, "utility": { "latency_ms": 0, "token_cost_factor": 0, "expected_success_rate": 0 }, "execution": { "uri": "${base_url}/v1/chat/completions", "model_id": "deepseek-v4-pro-0813" }, "memory": { "success_trajectories": [], "failure_reflections": [] }, "provenance": { "pricing_status": "official_public_price_snapshot", "pricing_note": "Official reference rates and backend billing rules from the same dated pricebook; capability and gateway evidence unchanged.", "supplier_model_mapping_status": "configured_from_user_screenshot_not_live_verified", "semantic_review": { "reviewed_at": "2026-09-09", "policy": "model-capabilities-v1", "source_urls": [ "https://api-docs.deepseek.com/quick_start/pricing/", "https://api-docs.deepseek.com/guides/thinking_mode/" ], "evidence_status": "provider_documented_not_gateway_verified", "source_document_sha256": { "https://api-docs.deepseek.com/quick_start/pricing/": "cf2c6fb2dd8a32a538f12a8176175b8809a3516326a5cb30dfe52d63c490a968", "https://api-docs.deepseek.com/guides/thinking_mode/": "f28c43248d26db1f27af0cb082abb00326c957d560d33a21839736edd1d10724" }, "task_profile_basis": "Task categories derived from documented operations; no empirical quality ranking or fixed framework role.", "context_policy": "Legacy configured limit is preserved separately from documented native context; neither proves the supplier limit." }, "model_distinction_review": { "evidence_kind": "provider_documented_design", "source_urls": [ "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/raw/main/README.md", "https://api-docs.deepseek.com/updates" ], "source_document_sha256": { "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/raw/main/README.md": "c4d714818a4d3333542edc7d38ea065825a0cf7aa8fea3605bbd1d1c18e4a610", "https://api-docs.deepseek.com/updates": "05b8f4da6a0b410ba575e543e5d0be9fbed07e9b5ea8d5555ece5455ec7627e9" }, "interpretation_limit": "Published task emphasis is non-exclusive. Architecture and product positioning do not establish supplier latency, measured quality, or a fixed framework role.", "policy": "model-distinctions-v1", "reviewed_at": "2026-09-09", "evidence_status": "provider_documented_not_gateway_verified", "scope": "Non-exclusive published task emphasis and architecture; no empirical cross-model quality ordering." }, "runtime_reliability_evidence": { "protocol": "sgar-model-runtime-reliability-v1", "endpoint_scope": "configured_openai_compatible_gateway", "reviewed_at": "2026-10-02", "short_bounded_response": { "status": "observed_success", "successful_call_count": 3, "maximum_observed_output_tokens": 93, "request_sha256s": [ "8556f131e1fa48b58c74c7b9d6454a4aeff8c399349b56709e93c259e68db167", "c5e5fbecd64f015acebaf525479355b2e96d6837463b99fcc7731c6054bad84c", "3f1724e496d0e841459de3ff09f8411034e65d904c71988ca612a1cf5528454c" ] }, "long_form_artifact_generation": { "status": "not_qualified", "no_response_timeout_count": 2, "attempt_timeout_seconds": 1200, "request_sha256s": [ "fc6e7e2535f9cdd537dab1870a445fe3fefabd497d4a7100fafd602c32e18c8f", "70d1594f5771bb4536b2d228a4694f464d35e3b42c7eeeafeb257ecaae77320e" ], "billing_status": "provider_usage_unknown" } }, "official_pricing": { "snapshot_date": "2026-10-09", "source": "https://api-docs.deepseek.com/quick_start/pricing/", "pricebook_sha256": "9e9a8248e54f1934d08d5efe68a3998ebce6faa09106f3f2a7ac119bf89a182c" } }, "type_specific": { "model": { "profile_version": "model-capabilities-v1", "input_modalities": [ "text" ], "output_modalities": [ "text" ], "supports": { "tool_calling": true, "json_mode": true, "vision": false, "reasoning": true, "coding": true }, "thinking_mode": "hybrid", "native_protocol": "openai_compatible_chat", "documented_context_tokens": 1048576, "evidence_status": "provider_documented_not_gateway_verified", "gateway_verification_status": "not_run", "limitations": [ "The current gateway is qualified for short bounded responses only. Two independent long-form artifact-generation requests reached the 1200-second deadline without a response. Do not select this route for long-form code or large artifact generation; select another ready model." ], "pricing_rules": { "input_per_m": "1.32", "cache_per_m": ".044", "output_per_m": "3.96", "source": "https://api-docs.deepseek.com/quick_start/pricing/", "cache_mode": "implicit", "input_tiers": [] } } } }