Ultra Tech https://ultratechmanuals.com/ Full-length guides to AI software and hardware: every fact from its source, every example actually run, every guide checked before it is published. en Ollama 0.40.4 lets you ask decision models questions straight from the terminal https://ultratechmanuals.com/news/ollama-0-40-4 https://ultratechmanuals.com/news/ollama-0-40-4 Wed, 14 Oct 2026 00:00:00 GMT Ollama 0.40.4 brings decision models to the command line: you can now use them directly from the terminal with `ollama run`, to ask a yes/no question, sort a piece of text into labels, or score it against a rubric. Ollama 0.40.3 pauses the background model upgrade that 0.40.2 brought in https://ultratechmanuals.com/news/ollama-0-40-3 https://ultratechmanuals.com/news/ollama-0-40-3 Sun, 11 Oct 2026 00:00:00 GMT Ollama 0.40.3 pauses the background model upgrade that arrived with 0.40.2, and the reason given is the speed of embedding and other requests. Ollama 0.35 adds decision models that answer with choices, probabilities and scores instead of text https://ultratechmanuals.com/news/ollama-0-35-decision-models https://ultratechmanuals.com/news/ollama-0-35-decision-models Fri, 09 Oct 2026 00:00:00 GMT Ollama 0.35.0 adds decision models, served on your own machine through `/v1/systemone`, an endpoint the release page says is based on TypeSafe's Jev API. Ollama 0.40 runs supported models on MLX by default on Apple Silicon https://ultratechmanuals.com/news/ollama-0-40 https://ultratechmanuals.com/news/ollama-0-40 Fri, 09 Oct 2026 00:00:00 GMT Ollama 0.40.0 changes the engine that runs your models on an Apple Silicon Mac: model architectures that the MLX runtime supports now run on MLX automatically. Ollama 0.40.2 upgrades your models in the background and keeps the originals as backups https://ultratechmanuals.com/news/ollama-0-40-2 https://ultratechmanuals.com/news/ollama-0-40-2 Fri, 09 Oct 2026 00:00:00 GMT Ollama 0.40.2 upgrades the models you downloaded with earlier versions in the background, the first time you run each one, for better performance and compatibility when they run on llama.cpp. OpenAI adds Ultrafast mode to GPT-6.1 Sol, at six times the Standard price and with EU data residency https://ultratechmanuals.com/news/openai-gpt-6-1-sol-ultrafast https://ultratechmanuals.com/news/openai-gpt-6-1-sol-ultrafast Fri, 09 Oct 2026 00:00:00 GMT OpenAI added Ultrafast mode for GPT-6.1 Sol to the Responses API on 8 October 2026. OpenAI cuts its API usage tiers from five to three — Build, Launch and Grow https://ultratechmanuals.com/news/openai-usage-tiers-three https://ultratechmanuals.com/news/openai-usage-tiers-three Fri, 09 Oct 2026 00:00:00 GMT OpenAI has simplified its API usage tiers from five to three — Build, Launch and Grow — in an entry its changelog dates 6 October 2026. OpenAI releases GPT-6.1 Sol, at GPT-6 Sol's input and output prices and half its cached-input price https://ultratechmanuals.com/news/openai-gpt-6-1-sol https://ultratechmanuals.com/news/openai-gpt-6-1-sol Fri, 09 Oct 2026 00:00:00 GMT OpenAI released GPT-6.1 Sol, model id `gpt-6.1-sol`, on 29 September 2026, for complex coding and professional work at a lower cost than GPT-6 Astra. OpenAI releases the Decisions API in beta, typed answers from gpt-6-luna on v1/decisions https://ultratechmanuals.com/news/openai-decisions-api https://ultratechmanuals.com/news/openai-decisions-api Fri, 09 Oct 2026 00:00:00 GMT OpenAI released the Decisions API in beta with gpt-6-luna, on a new endpoint, v1/decisions; its API changelog lists it on 6 October 2026. llama.cpp 0.6.0 adds GLM-5.3-Flash and Clef, a decision-model API in the server, and new Metal kernels https://ultratechmanuals.com/news/llamacpp-0-6-0 https://ultratechmanuals.com/news/llamacpp-0-6-0 Fri, 09 Oct 2026 00:00:00 GMT llama.cpp 0.6.0 introduces a new extended batch API, adds GLM-5.3-Flash, the Clef decision model and MTP speculative decoding for Qwen4Exp, gives the server a /v1/systemone API for decision models, rebuilds the Web UI around the Hugging Face Hub, adds flash-attention kernels for Metal and Vulkan, and moves ggml to v0.26.0. Mac or PC for Running Models Locally https://ultratechmanuals.com/guides/mac-or-pc-for-local-models https://ultratechmanuals.com/guides/mac-or-pc-for-local-models Sun, 27 Sep 2026 00:00:00 GMT List the models you want to run and say how much memory each one needs, with one file you run yourself. OpenAI fixes a bug that degraded image understanding in GPT-6 Sol and GPT-6 Luna https://ultratechmanuals.com/news/openai-gpt-6-image-encoding-fix https://ultratechmanuals.com/news/openai-gpt-6-image-encoding-fix Sun, 27 Sep 2026 00:00:00 GMT OpenAI has fixed a bug in image encoding that degraded image understanding in GPT-6 Sol and GPT-6 Luna, and its changelog dates the fix 25 September. OpenAI shuts down the Sora 2 video models and the Videos API, with no replacement https://ultratechmanuals.com/news/openai-sora-2-api-shut-down https://ultratechmanuals.com/news/openai-sora-2-api-shut-down Sun, 27 Sep 2026 00:00:00 GMT OpenAI has shut down the Sora 2 video models and the Videos API: since 24 September 2026 they are no longer available. What MCP Is https://ultratechmanuals.com/guides/what-mcp-is https://ultratechmanuals.com/guides/what-mcp-is Sun, 27 Sep 2026 00:00:00 GMT Write an MCP server that offers one tool, and send it a message by hand to see exactly what it answers. What Structured Output Is https://ultratechmanuals.com/guides/what-structured-output-is https://ultratechmanuals.com/guides/what-structured-output-is Sun, 27 Sep 2026 00:00:00 GMT Write a JSON Schema for the data you need, and send it with a request so the reply follows it. llama.cpp 0.5.0 lets the server listen on several addresses and reads sampling settings from the environment https://ultratechmanuals.com/news/llamacpp-0-5-0 https://ultratechmanuals.com/news/llamacpp-0-5-0 Sun, 27 Sep 2026 00:00:00 GMT llama.cpp 0.5.0 focuses on backend performance and correctness, broader model coverage, and more robust server and router operation. OpenAI releases GPT-6 Sol and GPT-6 Luna, and halves the price of both tiers https://ultratechmanuals.com/news/openai-gpt-6 https://ultratechmanuals.com/news/openai-gpt-6 Wed, 23 Sep 2026 00:00:00 GMT OpenAI's API changelog recorded on 22 September 2026 that it [released GPT-6 Sol, model id gpt-6-sol, and GPT-6 Luna, model id gpt-6-luna]. OpenAI lets an administrator decide who may create an API key https://ultratechmanuals.com/news/openai-api-key-governance https://ultratechmanuals.com/news/openai-api-key-governance Tue, 22 Sep 2026 00:00:00 GMT OpenAI's API changelog recorded on 15 September 2026 that it [added API key creation governance controls at the organization and project levels]. OpenAI makes Prompt Cache Diagnostics generally available https://ultratechmanuals.com/news/openai-prompt-cache-diagnostics https://ultratechmanuals.com/news/openai-prompt-cache-diagnostics Tue, 15 Sep 2026 00:00:00 GMT OpenAI's Prompt Cache Diagnostics is now generally available in the Responses API, for GPT-5.6 and later supported models. llama.cpp 0.4.1 adds three model architectures and replaces its memory flags with --load-mode https://ultratechmanuals.com/news/llamacpp-0-4-1 https://ultratechmanuals.com/news/llamacpp-0-4-1 Tue, 15 Sep 2026 00:00:00 GMT llama.cpp 0.4.1 adds Maple 20B-A1B, Tencent Hy 4 and Spark2.5 support, improves JSON schema handling, chat parsing, logging and server child-process management, and updates ggml to v0.24.0.