diff --git a/agent_cost_estimator/pricing.json b/agent_cost_estimator/pricing.json index 4dd20cd..76dc909 100644 --- a/agent_cost_estimator/pricing.json +++ b/agent_cost_estimator/pricing.json @@ -16,7 +16,38 @@ "output_price": 7.5, "note": "Global fallback rates for unknown providers/models", "source": "agent-cost-estimator default", - "date": "2026-11-10" + "date": "2026-11-18" + } + } + }, + "deepseek": { + "currency": "USD", + "provider_alias": "mistral", + "note": "DeepSeek models as hosted through Mistral's marketplace (Mistral API key, no separate DeepSeek account). Section is named deepseek because that is the provider/model prefix in harness logs; provider_alias makes the cost tracker display them as mistral:.", + "models": { + "default": { + "input_price": 0.3, + "cached_input_price": 0.007, + "output_price": 1.2, + "note": "DeepSeek default rates - matches V4.1 Flash 1017 (Mistral hosted)", + "source": "https://docs.mistral.ai/inference/pricing", + "date": "2026-11-18" + }, + "deepseek-v4-1-flash-1017": { + "input_price": 0.3, + "cached_input_price": 0.007, + "output_price": 1.2, + "note": "DeepSeek V4.1 Flash 1017 (Mistral-hosted model id deepseek-v4-1-flash-1017; 552B MoE, 1M context, MIT weights). DeepSeek's direct API serves the same release as deepseek-flash at $0.30 in / $0.006 cached / $1.20 out peak, half off-peak - see https://benchlm.ai/deepseek/api-pricing", + "source": "https://docs.mistral.ai/inference/pricing", + "date": "2026-11-18" + }, + "deepseek-flash": { + "input_price": 0.3, + "cached_input_price": 0.007, + "output_price": 1.2, + "note": "deepseek-flash alias (DeepSeek direct API name for the V4.1-Flash-1017 release) - same rates as the Mistral-hosted deepseek-v4-1-flash-1017", + "source": "https://docs.mistral.ai/inference/pricing", + "date": "2026-11-18" } } }, @@ -27,25 +58,25 @@ "input_price": 1.5, "cached_input_price": 0.15, "output_price": 7.5, - "note": "Mistral default rates - matches Mistral Medium 3.5", + "note": "Mistral default rates - matches Mistral Medium 3.5 (off the official pricing table since ~2026-11-17; unchanged rates confirmed by benchlm.ai and optimtoken.optimnow.io)", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-medium": { "input_price": 1.5, "cached_input_price": 0.15, "output_price": 7.5, - "note": "Mistral Medium 3.5 - flagship reasoning model", + "note": "Mistral Medium 3.5 - flagship reasoning model; no longer listed on the official pricing table (gone by 2026-11-17) - unchanged rates confirmed by benchlm.ai and optimtoken.optimnow.io", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-medium-3.5": { "input_price": 1.5, "cached_input_price": 0.15, "output_price": 7.5, - "note": "Vibe's built-in hosted default model rates", + "note": "Vibe's built-in hosted default model; no longer listed on the official pricing table (gone by 2026-11-17) - unchanged rates confirmed by benchlm.ai and optimtoken.optimnow.io", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-small": { "input_price": 0.15, @@ -53,7 +84,7 @@ "output_price": 0.6, "note": "Mistral Small 4 model rates", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-small-4": { "input_price": 0.15, @@ -61,7 +92,7 @@ "output_price": 0.6, "note": "Mistral Small 4 - efficient reasoning model", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-small-latest": { "input_price": 0.15, @@ -69,15 +100,15 @@ "output_price": 0.6, "note": "Mistral Small -latest alias served by the API (same rates as Mistral Small 4)", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-11" + "date": "2026-11-18" }, "mistral-large": { "input_price": 0.7, "cached_input_price": 0.07, "output_price": 2.1, - "note": "Mistral Large 4 rates (alias of mistral-large-4); promotional rates as of 2026-11-10", + "note": "Mistral Large 4 rates (alias of mistral-large-4); promotional rates still listed 2026-11-18", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-large-3": { "input_price": 0.5, @@ -85,23 +116,23 @@ "output_price": 1.5, "note": "Mistral Large 3 - previous generation large model (no longer listed on the official pricing page)", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-large-4": { "input_price": 0.7, "cached_input_price": 0.07, "output_price": 2.1, - "note": "Mistral Large 4 - natively multimodal MoE (1.05T params); promotional rates as of 2026-11-10 (regular: $1.36 in / $0.14 cached / $4.18 out)", + "note": "Mistral Large 4 - natively multimodal MoE (1.05T params); promotional rates still listed 2026-11-18 (regular: $1.36 in / $0.14 cached / $4.18 out)", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "mistral-vibe-cli-latest": { "input_price": 1.5, "cached_input_price": 0.15, "output_price": 7.5, - "note": "Mistral Vibe CLI latest model (default hosted model - Mistral Medium 3.5)", + "note": "Mistral Vibe CLI latest model (default hosted model - Mistral Medium 3.5; off the official pricing table since ~2026-11-17, unchanged rates confirmed by benchlm.ai)", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "ministral-3-3b": { "input_price": 0.1, @@ -109,7 +140,7 @@ "output_price": 0.1, "note": "Ministral 3 3B - compact edge model with symmetric pricing", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "ministral-3-8b": { "input_price": 0.15, @@ -117,7 +148,7 @@ "output_price": 0.15, "note": "Ministral 3 8B - edge model with symmetric pricing", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "ministral-3-14b": { "input_price": 0.2, @@ -125,7 +156,7 @@ "output_price": 0.2, "note": "Ministral 3 14B - edge model with symmetric pricing", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "codestral": { "input_price": 0.3, @@ -133,7 +164,7 @@ "output_price": 0.9, "note": "Codestral - specialized code generation model", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" } } }, @@ -148,7 +179,7 @@ "output_price": 4.4, "note": "Z.ai default rates - matches GLM 5.3 (Mistral hosted)", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "zai-glm-5-3": { "input_price": 1.4, @@ -156,7 +187,7 @@ "output_price": 4.4, "note": "GLM 5.3 (Mistral hosted; billed via Mistral's marketplace as zai-glm-5-3). Z.ai direct API lists cached input at $0.26/1M - see https://docs.z.ai/guides/overview/pricing", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "zai-glm-5.3-flash": { "input_price": 0.15, @@ -164,15 +195,15 @@ "output_price": 0.5, "note": "GLM 5.3 Flash (Mistral hosted) - efficient MoE model (320B total, 18B active); same rates on Z.ai direct", "source": "https://docs.mistral.ai/inference/pricing", - "date": "2026-11-10" + "date": "2026-11-18" }, "zai-glm-5.3-flashx": { "input_price": 0.37, "cached_input_price": 0.075, "output_price": 1.25, - "note": "GLM 5.3 FlashX - advanced variant; not listed on Mistral's pricing page, rates are Z.ai direct (would differ if Mistral hosts it later)", + "note": "GLM 5.3 FlashX - advanced variant; not listed on Mistral's pricing page, rates are Z.ai direct (would differ if Mistral hosts it later); re-verified 2026-11-18 on docs.z.ai", "source": "https://docs.z.ai/guides/overview/pricing", - "date": "2026-11-10" + "date": "2026-11-18" } } }