Cool MiniMax-H3 use case unlocked for game sprites creation, here's the workflow: 1. one character design still, on a flat magenta background 2. one 5s H3 clip per move (walk, jump, punch, kick, block...) 3. cut each clip down to its pose extremes 4. key out the magenta, pack into a sprite atlas wire that to a state machine to make it playable (8 moves, ~$5)

A coding agent just added support for itself to a 30k-star open-source project. MiniMax Code entered an unfamiliar Rust codebase, implemented its own state detection, passed 3,475 tests, and shipped a clean patch. Now Herdr knows when mcode is idle, working, or blocked. Discussion： Feat Branch：

I founded DeerFlow. I lead R&D for MiniMax Code. Today, the two systems officially crossed wires. 🦌⚡ Thanks to @willemjiang @henry19840301 MiniMax Code now runs inside DeerFlow as a native ACP agent: zero adapter isolated workspaces MCP passthrough clean agent-to-agent output One command: mcode acp The agent stack is starting to compose itself.

Here is how to enable a 1M-token context window in Codex for GPT-5.6 Sol. Even though we have tuned the context limit in Codex to be set optimally when it comes to performance and cost, this is a common ask, so here it is documented. A larger context window lets Codex retain more code, tool output, and conversation history before summarizing older material. You need a model that supports it. And GPT-5.6 Sol, for example, has a documented 1,050,000-token window. Open ~/.codex/config.toml and add or update these settings at the top level, before any [section] headers: ``` model = "gpt-5.6-sol" model_context_window = 1000000 model_auto_compact_token_limit = 900000 ``` The first setting selects the model. The second tells Codex to use a one-million-token context budget. The third starts automatic history compaction around 900,000 tokens, leaving some headroom. Restart Codex client and start a new session after saving. To try the configuration for a single CLI session without changing your defaults: ``` codex -m gpt-5.6-sol \ -c model_context_window=1000000 \ -c model_auto_compact_token_limit=900000 ``` Have fun, but also know that we tuned the default carefully!

With DeepSeek Harness bringing fresh attention to the space, one thing feels clear: the defining test of a harness is whether it can evolve with its developers. That loop is already taking shape around MCode CLI. Developers are using MCode to evolve MCode itself—building everything from a Chinese localization extension to a gloriously unhinged token burner. Their creativity has blown us away. Next week, we’re opening the official MCode CLI Extensions repository and inviting every developer to step into the loop—and help shape what MCode becomes next. Stay tuned.

hello, world! i'm daniel! currently i'm the head of r&d of MiniMax code @MiniMax_AI, which is a frontier ai lab in china. i enjoy working on our ai-native product @MiniMaxAgent . i'm really obsessed with this question: what does an ai-native organization look like? how can our organization co-evolve together with llms? besides that, i'm passionate about open source. i'm the co-author of the deerflow project, which is a representative work in harness engineering and has 78k 🌟 on github. if you're passionate about open source, foundation models, or agents- or if you enjoy playing basketball or riding a road bike - let's chat and be friends! :)