🪢 Open source agent evals & observability: Trace, evaluate, and improve LLM applications with one open platform.
TypeScript analyticsautogenevaluationlangchainlarge-language-modelsllama-indexllmllm-evaluationllm-observabilityllmopsmonitoringobservabilityopen-sourceopenaiplaygroundprompt-engineeringprompt-managementself-hostedycombinator
chore(pricing): add OpenAI gpt-6-astra default pricing (#17030)
langfuse-botfbc2c07b1cf2
diff --git a/.agents/skills/add-model-price/references/provider-sources-and-price-keys.md b/.agents/skills/add-model-price/references/provider-sources-and-price-keys.mdindex 19477ac92a..c68212ddc0 100644--- a/.agents/skills/add-model-price/references/provider-sources-and-price-keys.md+++ b/.agents/skills/add-model-price/references/provider-sources-and-price-keys.md@@ -459,6 +459,41 @@ file and `openAIModels`in July 27 2026 audit. Official sources:`https://developea third model sharing the same Jan 1, 2027 price step-up — see unresolved finding in`model-audit-memory.md` about updating all three (3.6, 3.7, 3.8 Flash) on/after thatdate.+- **GPT-6 Astra (added September 3 2026)** — `gpt-6-astra` is OpenAI's new+ flagship reasoning model ("Our most capable model, built for the hardest+ end-to-end work"), confirmed via `https://developers.openai.com/api/docs/pricing`+ and its dedicated model page `https://developers.openai.com/api/docs/models/gpt-6-astra`.+ API ID: `gpt-6-astra` (no date-stamped snapshot at launch, following the+ gpt-5.6 family's dateless-launch precedent). Context window 1,050,000 tokens+ (max input 922,000), max output 128,000 tokens, prompt caching and reasoning+ tokens supported. Standard pricing: $10/MTok input, $1/MTok cached input,+ $12.50/MTok cache write (1.25x input, matching the gpt-5.6-family pattern),+ $50/MTok output. **Large Context tier applies above 272,000 input tokens**+ (the same threshold as the gpt-5.6 family, not the 200K used by most other+ OpenAI models): 2x input/cache prices and 1.5x output for the full request —+ $20/$2/$25/$75. Fast mode (`service_tier` in `["fast","priority"]`) is 2x the+ applicable tier's rate: $20/$2/$25/$100 standard, $40/$4/$50/$150 large+ context (Fast mode is documented as unavailable for EU data residency+ requests — Langfuse's pricing schema cannot condition on data residency, so+ the tier is added unconditionally like every other Fast-mode tier). Flex+ (`service_tier: "flex"`) is 0.5x the applicable tier's rate: $5/$0.50/$6.25/$25+ standard, $10/$1/$12.50/$37.50 large context. Added to the pricing file+ mirroring the `gpt-5.6-sol` six-tier key set exactly (Standard, Fast mode ·+ Large context, Flex · Large context, Fast mode, Flex, Large Context) and to+ `openAIModels` in `types.ts` (not as the first entry). matchPattern:+ `(?i)^(openai/)?(gpt-6-astra)$`. Confirmed via two independent WebFetch+ calls (aggregate pricing table plus the dedicated model page) that returned+ consistent numbers. **Batch pricing intentionally not added**: the pricing+ page also documents a Batch tier at 50% of Standard (same discount as Flex),+ but no OpenAI model in this file has ever had a Batch tier represented —+ Batch is a distinct async submission endpoint (results collected up to 24h+ later via a separate Batches API) rather than a `service_tier` value on a+ normal chat/response request, so it would not appear in ordinary Langfuse+ ingestion usage data the way `fast`/`priority`/`flex` do. Treat this as the+ established scope boundary rather than a gap to fill without first+ confirming Langfuse actually observes a `service_tier: "batch"` (or+ equivalent) value in real ingested usage payloads for any provider request+ path.- **Gemini 2.5-family grounding price is a different rate than 3.x — confirmed thepricing file correctly has no grounding keys on 2.5-family models (September 22026)** — `gemini-2.5-pro`/`gemini-2.5-flash`/`gemini-2.5-flash-lite` grounding isdiff --git a/packages/shared/src/server/llm/types.ts b/packages/shared/src/server/llm/types.tsindex 4d2ee726fa..bc1a12ccf7 100644--- a/packages/shared/src/server/llm/types.ts+++ b/packages/shared/src/server/llm/types.ts@@ -320,6 +320,7 @@ export const openAIModels = ["gpt-4.1-mini-2025-04-14","gpt-4.1-nano","gpt-4.1-nano-2025-04-14",+ "gpt-6-astra","gpt-5.6-sol","gpt-5.6-terra","gpt-5.6-luna",diff --git a/worker/src/constants/default-model-prices.json b/worker/src/constants/default-model-prices.jsonindex f0f1574d47..922ddd9194 100644--- a/worker/src/constants/default-model-prices.json+++ b/worker/src/constants/default-model-prices.json@@ -6405,6 +6405,182 @@}]},+ {+ "id": "92f817ff-d8a1-479b-87eb-405f9c7b4a18",+ "modelName": "gpt-6-astra",+ "matchPattern": "(?i)^(openai/)?(gpt-6-astra)$",+ "createdAt": "2026-09-03T00:00:00.000Z",+ "updatedAt": "2026-09-03T00:00:00.000Z",+ "tokenizerConfig": {+ "tokensPerName": 1,+ "tokenizerModel": "gpt-4",+ "tokensPerMessage": 3+ },+ "tokenizerId": "openai",+ "pricingTiers": [+ {+ "id": "92f817ff-d8a1-479b-87eb-405f9c7b4a18_tier_default",+ "name": "Standard",+ "isDefault": true,+ "priority": 0,+ "conditions": [],+ "prices": {+ "input": 10e-6,+ "input_cached_tokens": 1e-6,+ "input_cache_read": 1e-6,+ "cache_read_input_tokens": 1e-6,+ "input_cache_creation": 12.5e-6,+ "cache_write_tokens": 12.5e-6,+ "output": 50e-6,+ "output_reasoning_tokens": 50e-6,+ "output_reasoning": 50e-6,+ "reasoning_tokens": 50e-6+ }+ },+ {+ "id": "d3151184-4443-430a-a4ed-961dd55d223c",+ "name": "Fast mode · Large context (>272K)",+ "isDefault": false,+ "priority": 1,+ "conditions": [+ {+ "source": "model_parameters",+ "key": "service_tier",+ "operator": "in",+ "values": ["fast", "priority"]+ },+ {+ "usageDetailPattern": "(input|cache_write)",+ "operator": "gt",+ "value": 272000,+ "caseSensitive": false+ }+ ],+ "prices": {+ "input": 40e-6,+ "input_cached_tokens": 4e-6,+ "input_cache_read": 4e-6,+ "cache_read_input_tokens": 4e-6,+ "input_cache_creation": 50e-6,+ "cache_write_tokens": 50e-6,+ "output": 150e-6,+ "output_reasoning_tokens": 150e-6,+ "output_reasoning": 150e-6,+ "reasoning_tokens": 150e-6+ }+ },+ {+ "id": "1b8d0945-bd74-4223-8bed-320f2063a2db",+ "name": "Flex · Large context (>272K)",+ "isDefault": false,+ "priority": 2,+ "conditions": [+ {+ "source": "model_parameters",+ "key": "service_tier",+ "operator": "in",+ "values": ["flex"]+ },+ {+ "usageDetailPattern": "(input|cache_write)",+ "operator": "gt",+ "value": 272000,+ "caseSensitive": false+ }+ ],+ "prices": {+ "input": 10e-6,+ "input_cached_tokens": 1e-6,+ "input_cache_read": 1e-6,+ "cache_read_input_tokens": 1e-6,+ "input_cache_creation": 12.5e-6,+ "cache_write_tokens": 12.5e-6,+ "output": 37.5e-6,+ "output_reasoning_tokens": 37.5e-6,+ "output_reasoning": 37.5e-6,+ "reasoning_tokens": 37.5e-6+ }+ },+ {+ "id": "2da4af90-9c52-452a-b9d2-73e8306bad6e",+ "name": "Fast mode",+ "isDefault": false,+ "priority": 3,+ "conditions": [+ {+ "source": "model_parameters",+ "key": "service_tier",+ "operator": "in",+ "values": ["fast", "priority"]+ }+ ],+ "prices": {+ "input": 20e-6,+ "input_cached_tokens": 2e-6,+ "input_cache_read": 2e-6,+ "cache_read_input_tokens": 2e-6,+ "input_cache_creation": 25e-6,+ "cache_write_tokens": 25e-6,+ "output": 100e-6,+ "output_reasoning_tokens": 100e-6,+ "output_reasoning": 100e-6,+ "reasoning_tokens": 100e-6+ }+ },+ {+ "id": "ec8b3a8e-86c2-4a31-8c60-5f8c4ade98a8",+ "name": "Flex",+ "isDefault": false,+ "priority": 4,+ "conditions": [+ {+ "source": "model_parameters",+ "key": "service_tier",+ "operator": "in",+ "values": ["flex"]+ }+ ],+ "prices": {+ "input": 5e-6,+ "input_cached_tokens": 0.5e-6,+ "input_cache_read": 0.5e-6,+ "cache_read_input_tokens": 0.5e-6,+ "input_cache_creation": 6.25e-6,+ "cache_write_tokens": 6.25e-6,+ "output": 25e-6,+ "output_reasoning_tokens": 25e-6,+ "output_reasoning": 25e-6,+ "reasoning_tokens": 25e-6+ }+ },+ {+ "id": "296fbd59-2e9a-43c0-8308-cbc3b9014fdd",+ "name": "Large Context (>272K)",+ "isDefault": false,+ "priority": 5,+ "conditions": [+ {+ "usageDetailPattern": "(input|cache_write)",+ "operator": "gt",+ "value": 272000,+ "caseSensitive": false+ }+ ],+ "prices": {+ "input": 20e-6,+ "input_cached_tokens": 2e-6,+ "input_cache_read": 2e-6,+ "cache_read_input_tokens": 2e-6,+ "input_cache_creation": 25e-6,+ "cache_write_tokens": 25e-6,+ "output": 75e-6,+ "output_reasoning_tokens": 75e-6,+ "output_reasoning": 75e-6,+ "reasoning_tokens": 75e-6+ }+ }+ ]+ },{"id": "6ed696e7-5f3a-4113-a233-019155fb3ff0","modelName": "gemini-3.5-flash",