GPT-5 and o3 API shutdown, Dec 11 2026: how to migrate to GPT-5.6 Sol, Terra, Luna
OpenAI retires GPT-5, mini, nano, o3 and Pro on Dec 11, 2026. Move to GPT-5.6 Sol, Terra or Luna: input costs 2-8x more; retest tools and effort.
Read postAIVAX / Blog
Notes from the operating layer.
Field notes on AI infrastructure, retrieval, model operations and the systems that keep every request visible.
32 published field notesLatest blog posts
-
-
GPT-4o and GPT-4.1 API shutdown on January 14, 2027: how to migrate to gpt-6-luna
OpenAI removes GPT-4o and GPT-4.1 from the API on January 14, 2027. Migrate to gpt-6-luna: effort none, no temperature/top_p, re-test tools and cost.
Read post -
How to fix 'Function tools with reasoning_effort are not supported' in Chat Completions
Fix the 400 on gpt-5.6-sol Chat Completions: set reasoning_effort to none or move function tools to Responses. Includes AIVAX BYOK setup.
Read post -
How to filter RAG results by metadata, tags and date
Filter RAG results by metadata, tags and date before ranking, or run a filter-only search. Syntax, common mistakes and a working AIVAX request.
Read post -
Why is my prompt cache hit rate low? Prefix breakers in OpenAI and Anthropic requests
Prompt caches reuse only an identical prefix. Check timestamps, tools, compaction, and TTL, and compute the hit rate correctly, including on AIVAX.
Read post -
OpenAI Evals shutdown: where do multi-turn agent tests go before November 30?
Manually recreate OpenAI Evals datasets and graders in Promptfoo, then choose between its multi-turn tests and AIVAX gateway conversations.
Read post -
Too many MCP tools: how to reduce tool-definition tokens in your agent's context
Measure what MCP tool definitions cost, then cut it with allowlists, deferred tool search, or a shell. What each option needs and what AIVAX supports.
Read post -
How to convert a PDF to Markdown for RAG and split it into chunks
Convert a PDF to Markdown with OCR, split it into chunks for retrieval, and export JSONL for a RAG collection. Code and a free demo included.
Read post -
OpenAI GPT-4, o1, o3-mini and o4-mini shutdown on October 23, 2026: what breaks and what to check
OpenAI shut down GPT-4, o1, o3-mini, o4-mini and other models on October 23, 2026, so requests naming them now fail. Find the IDs left in fallbacks and jobs, then check each replacement's parameters, output and cost.
Read post -
How to generate images and speech from Claude Code, Cursor or VS Code with MCP
Configure AIVAX's hosted Media Generation MCP, discover models and prices, and generate images or MP3 speech with explicit tool and privacy limits.
Read post -
How to make LLM tool calls idempotent so a retry doesn't create a duplicate ticket
Deduplicate the same tool operation; changed arguments create a new one. Require an explicit conflict policy to allow only one write per request.
Read post -
How to check PDF and scan ingestion quality before you embed for RAG
Before embedding, sample real pages, write the facts you expect, and grade each generated document for wrong numbers, lost table headers, and missing facts.
Read post -
How MCP skills carry a workflow across the server boundary
SEP-2640 lets MCP servers offer discoverable workflow skills as resources. The host still decides what to load, verify, approve, and execute.
Read post -
Free RAG API: embeddings, reranking and OCR limits on AIVAX Free
AIVAX Free has separate daily allowances for RAG embeddings, Reflex reranking, Julia-1 decisions and Fetch/OCR. What each covers and what is still billed.
Read post -
How to fix invalid JSON from an LLM API with structured outputs
Check truncation, refusals and schema support first. In AIVAX, distinguish native response_format from response_schema validation and bounded healing.
Read post -
Route by complexity, price by the task
A lower token rate does not tell you what a completed task costs. AIVAX's Complexity Router selects a model and reasoning effort for each request so teams can measure quality and spend at the task level.
Read post -
How to extract text from a URL, scanned PDF or image for an LLM pipeline
Use AIVAX Fetch to extract text from URLs, scanned PDFs and images, with per-item errors, processing-unit usage and optional schema-guided JSON.
Read post -
Decisions API example: triage a support ticket by urgency, department and frustration
Triage a support ticket in one Decisions API call: urgency (noul), team (choice) and frustration (score), with curl, routing code and model prices.
Read post -
Web search API vs. page fetch for LLM research: why snippets are not enough
Use search to discover URLs and fetch to read their content. Retain queries, results, and extracted passages so LLM answers can cite evidence.
Read post -
Give your coding agent a memory it can read and a world it can fetch
Coding agents lose context between sessions and guess at the world outside the repo. AIVAX's Collections and Web Utilities MCP servers give them writable semantic memory plus fetch and search — with scoped credentials, bounded retrieval, and explicit write control.
Read post -
How to protect LLM agent memory from poisoning
Secure persistent agent memory with scoped writes, expiry, provenance, and review. Understand AIVAX externalUserId boundaries and deletion limits.
Read post -
How to run thousands of LLM requests in batch with rate limits and retries
Use bounded concurrency, balance checks, targeted retries, and resumable jobs to process thousands of LLM requests without replaying completed work.
Read post -
From test to monitor: the judge trajectory as a production signal
Agentic Tests already scores every turn of a simulated conversation. Applied to real production traffic, the same trajectory signal — score, at-risk state, persistent loss — tells you when a live conversation is drifting before the user gives up.
Read post -
Migrate an MCP server off Mcp-Session-Id: the stateless 2026-07-28 spec
MCP 2026-07-28 drops initialize and Mcp-Session-Id. Move state to explicit handles, make retries safe, and keep a legacy path for older clients.
Read post -
RAG vs vector database: is vector search enough?
A vector database handles vector search. RAG also needs source preparation, updates and context assembly. Compare what to build and what to manage.
Read post -
How to safely connect third-party MCP servers
Review MCP tools, credentials and imported instructions before connecting. Apply authorization and approval rules, with AIVAX configuration examples.
Read post -
Migrate OpenAI Assistants threads, runs and file search to the Responses API
Map Assistants threads, runs and file search to Responses after the shutdown. Recover saved history and decide which state your application should own.
Read post -
How to set reasoning effort across providers: OpenAI, Anthropic and Gemini
Compare OpenAI reasoning effort, Anthropic thinking and Gemini thinking controls. Configure AIVAX requests and handle summaries, state and token costs.
Read post -
LLM input moderation: system prompt or separate moderation step?
Use a separate moderation step to decide whether input may proceed before generation. System prompts guide behavior; they do not enforce permissions.
Read post -
How to test multi-turn AI agent conversations with a simulated user and an LLM judge
Define an observable goal, simulate a user, and judge the full conversation. Inspect weak turns, recovery, outcomes, and costs with AIVAX Agentic Tests.
Read post -
Do I Need a Reranker for RAG? Semantic Search vs. Reranking
Test a reranker when relevant documents rank too low. Fix missing candidates first, then measure answer quality, latency, and cost.
Read post -
Rerank RAG Results Without a Cross-Encoder: When to Use Reflex
Reflex reranks RAG candidates without a cross-encoder, using semantic and lexical signals and per-account reuse of document representations.
Read post