AICoder

AI Coding Live News

First-hand AI coding updates from 𝕏/Twitter frontier labs and researchers, distilled by AI with entity benchmarks.

Explore live on 𝕏 ↗

AI Coding Live News

K
Kimi.ai@Kimi_Moonshot·
👀156
🔥90
View on 𝕏 ↗

Moonshot launches Kimi K3.1 Pro: 2.8T-total / 104B-active open-weight MoE with 1M context, 65.7% on Terminal-Bench 4.0, API at $2/$10, weights by Nov 4

On 2026-10-23 (17:15 UTC) the official @Kimi_Moonshot account launched Kimi K3.1 Pro, alongside K3.1 Pro Highspeed (2-3x faster). Both are live on Kimi.ai, Kimi Work, Kimi Code and the Kimi API, and full weights are promised by 2026-11-04. The official tech blog describes it as a post-trained version of K3 with much larger multi-domain joint RL: a 2.8T-total / 104B-active MoE, 1M-token context, multimodal. Vendor-reported scores: 65.7% on Terminal-Bench 4.0 (GPT-6 Astra 58.2%, Claude Opus 5.5 66.4%) and 92.7% on BenchCAD with tools. The Kimi API pricing page lists kimi-k3.1-pro at $2 input, $10 output and $0.20 cached input per million tokens, down from K3 at $3/$15.

Moonshot launches Kimi K3.1 Pro: 2.8T-total / 104B-active open-weight MoE with 1M context, 65.7% on Terminal-Bench 4.0, API at $2/$10, weights by Nov 4

Key Takeaways

  • 2.8T-total / 104B-active MoE, 1M-token context, multimodal; full weights promised by 2026-11-04 (official blog)
  • Vendor-reported: 65.7% on Terminal-Bench 4.0, ahead of GPT-6 Astra (58.2%) and just behind Claude Opus 5.5 (66.4%); 61.2% on the internal Kimi Code Bench 3.0 Long-horizon (K3: 55.95%)
  • API pricing: kimi-k3.1-pro at $2 input / $10 output / $0.20 cached input per million tokens; Highspeed at $4 / $20 (pricing page)
  • On the internal Kimi Hallu Bench Hard stress test, claim-level hallucination fell from 23.4% (K3) to 8.3%; Moonshot notes this is not a daily-use error rate
  • Cyber-related data was removed from mid-training and RL for the open-weight release: xdev-bench-hard pass@3 drops from 90.5% (unmitigated checkpoint) to 81.0%
Read full story
Linked Models & Agents:
⚡ Kimi K3.1 Pro⚡ Kimi K3.1 Pro Highspeed⚡ Kimi K3🤖 Kimi Code🌐 Moonshot AI
#Kimi#Moonshot AI#open weights#MoE#1M context#coding agent#API pricing
S
Stefano Ermon@StefanoErmon·
👀70
🔥95
View on 𝕏 ↗

Inception Unveils Mercury 3: Diffusion-Based LLM Breaks 1,000+ Tok/s with 1M Context Window

Inception co-founder Stefano Ermon officially unveiled Mercury 3, the company's next-generation diffusion-based large language model. Delivering an astonishing 1,068 tokens per second inference speed on NVIDIA GPUs, Mercury 3 scales to 40x the size of Mercury 1 with a native 1M context window. Benchmarked on IFBench, GPQA Diamond, and SWE-Bench Atlas, it matches frontier closed baselines with standard API pricing of $0.20 / $0.70 per 1M input/output tokens (with 50% promotional pricing of $0.10 / $0.35).

Inception Unveils Mercury 3: Diffusion-Based LLM Breaks 1,000+ Tok/s with 1M Context Window

Key Takeaways

  • Inception co-founder Stefano Ermon officially launched Mercury 3, a diffusion LLM generating 1,068 tokens per second on NVIDIA GPUs
  • Scales model size by 40x over Mercury 1 with a native 1M context window; priced at $0.20 / $0.70 per 1M tokens ($0.10/$0.35 promotional)
  • Triples baseline intelligence on IFBench, GPQA Diamond, and SWE-Bench Atlas, with live availability via OpenRouter
  • Bypasses autoregressive sequential bottlenecks via parallel diffusion denoising, accelerating multi-step agent workflows
Read full story
Linked Models & Agents:
#Inception#Mercury 3#Diffusion LLM#1000 TPS#Stefano Ermon#OpenRouter
I
Inception_inception_ai·
👀242
🔥92
View on 𝕏 ↗

Inception launches Mercury 3 diffusion LLM: 1,068 tokens/s and 1M context, beating GPT-6 Luna and Haiku 5.5 on several benchmarks per the vendor

On October 19 Inception released Mercury 3, which it calls the most capable diffusion LLM (dLLM) to date. Per the vendor, it is more than 40x the size of Mercury 1, outputs 1,068 tokens/s, expands context from 260K (Mercury 2.5) to 1M tokens, and beats GPT-6 Luna and Claude Haiku 5.5 (both medium) on GPQA Diamond, IFBench, SWE-Bench Atlas and more. List price is $0.20/M input and $0.70/M output, 50% off at launch. It is live on the Inception API, OpenRouter, Baseten and Vercel AI Gateway with an OpenAI-compatible API. All scores are vendor-reported and not yet independently reproduced.

Inception launches Mercury 3 diffusion LLM: 1,068 tokens/s and 1M context, beating GPT-6 Luna and Haiku 5.5 on several benchmarks per the vendor

Key Takeaways

  • Speed: 1,068 tokens/s output; the vendor table lists GPT-6 Luna at ~130 and Haiku 5.5 at ~240 (vendor figures).
  • Scores: GPQA Diamond 89.23, IFBench 76.84, SWE-Bench Atlas 51.08, all above Luna and Haiku 5.5; but DeepSWE v1 at 46.31 trails Haiku 5.5 (51.62), and MRCR v2 also trails Haiku.
  • Scale and context: more than 40x the parameters of Mercury 1, with a 1M-token window (260K on Mercury 2.5).
  • Price: $0.20/M in and $0.70/M out list, 50% off at launch ($0.10/$0.35), cached input 90% off; new API keys get 100M free tokens.
  • Access: Inception API, OpenRouter, Baseten and Vercel AI Gateway, OpenAI-compatible as a drop-in replacement.
Read full story
Linked Models & Agents:
⚡ Mercury 3⚡ Mercury 2.5⚡ GPT-6 Luna⚡ Claude Haiku 5.5🌐 Inception🌐 OpenRouter🌐 Baseten🌐 Vercel
#Diffusion LLM#Mercury#Inception
M
MiniMax@MiniMax_AI·
👀98
🔥95
View on 𝕏 ↗

MiniMax Releases M3.1 Flash: 1M Context Window, $0.15/M Token Pricing, Surging +106 Points in WebDev Code Arena

MiniMax officially announced M3.1 Flash, its next-generation lightweight frontier model (previously dominating leaderboards as the secret model Space Bunny across 200T+ tokens). Featuring a native 1M context window, flat-rate API pricing of $0.15 / $0.60 per 1M input/output tokens ($0.015 cache reads), and 3x faster inference at 1/3 the cost of M3. It surged +106 points to 1588 on LMSYS Code Arena: WebDev (#12 among open models), with full weights set for open release on October 23.

MiniMax Releases M3.1 Flash: 1M Context Window, $0.15/M Token Pricing, Surging +106 Points in WebDev Code Arena

Key Takeaways

  • MiniMax officially launched M3.1 Flash, unveiling that it was the mystery model Space Bunny evaluated across 200T+ tokens
  • Features native 1M context with flat-rate API pricing of $0.15/$0.60 per 1M tokens ($0.015 cache reads), offering 3x speed at 1/3 cost of M3
  • Debuts on LMSYS Code Arena: WebDev at 1588 Elo (+106 pts over M3), ranking #12 among open models with major leaps in Gaming (+179) and Reference Design (+155)
  • Artificial Analysis benchmark confirms Intelligence Index of 42 and HLE of 48% at half the token consumption of rivals; weights release on Oct 23
Read full story
Linked Models & Agents:
#MiniMax#M3.1 Flash#Space Bunny#Code Arena#WebDev#Artificial Analysis
ADSponsored
M
MiniMax (official)MiniMax_AI·
👀101
🔥92
View on 𝕏 ↗

MiniMax M3.1 Flash weights now open on Hugging Face: 428B-total / 23B-active MoE, $0.15 input and $0.60 output per million tokens, 73.7% on DeepSWE

MiniMax confirmed that Space Bunny, the anonymous model that led usage rankings on OpenRouter and elsewhere, was an early version of M3.1 Flash, and launched it on its API on October 19. Pay-as-you-go pricing is $0.15 input, $0.60 output and $0.015 cache read per million tokens, with a 1M context, text, image and video input, and an effort setting for thinking depth. Per MiniMax, it costs one third of M3 and runs 3x faster, scoring 73.7% on DeepSWE and 62.2% on Automation-Bench; the weights went live on Hugging Face on schedule on October 23 (MiniMax Community license, MXFP8), with day-0 support from Together, Inferact (vLLM) and LightSeek TokenSpeed.

MiniMax M3.1 Flash weights now open on Hugging Face: 428B-total / 23B-active MoE, $0.15 input and $0.60 output per million tokens, 73.7% on DeepSWE

Key Takeaways

  • Official pay-as-you-go pricing: $0.15 input / $0.60 output / $0.015 cache read per million tokens; per token that is half of M3 ($0.30/$1.20), and MiniMax does not explain how it gets one third of the cost.
  • Vendor benchmarks: DeepSWE v1.1 73.7% (Opus 5: 74.0%), Automation-Bench 62.2%, BabyVision with tools 91.0%, Terminal Bench 4.0 38.1% (Opus 5: 51.8%).
  • 1M context, text/image/video input, five effort levels from low to max, default max, thinking cannot be turned off.
  • Weights opened October 23 at Hugging Face MiniMaxAI/MiniMax-M3.1-Flash: ungated, MiniMax Community license (not MIT/Apache), MXFP8 quantized, native multimodal MoE with about 428B total and 23B active parameters per token, plus a DSpark speculative-decoding draft head.
  • Migration: M2.5 and M2.7 closed to new users on October 19 and retire on November 18; M Plan and Token Plan requests now route to M3.1 Flash automatically.
  • Day-0 serving: Together AI runs it fully managed, MiniMax is partnering with Inferact to contribute production-inference improvements back to vLLM, and LightSeek's TokenSpeed added day-0 support.
Read full story
Linked Models & Agents:
⚡ MiniMax-M3.1-Flash⚡ MiniMax-M3⚡ DeepSeek-V4.1-Flash⚡ Kimi K3⚡ GLM-5.3⚡ GPT-5.6 Sol⚡ Claude Opus 5🤖 MiniMax Code🌐 MiniMax
#MiniMax#M3.1 Flash#Space Bunny#AI Coding#Agent#1M Context
@
@GoogleDevs@GoogleDevs·
👀67
🔥95
View on 𝕏 ↗

Google AI Studio Opens Gemini 4 Argon in Build & Playground for Pro Users; Transformers v5.20.1 Adds Native Reflection AI Beam and Meta 816B Muse Spark Open

Google AI Studio launched Gemini 4 Argon access in AI Studio Build and Playground for Google AI Pro subscribers, while Hugging Face Transformers v5.20.0 and v5.20.1 added native support for Reflection AI's frontier coding model Beam, Meta's 816B-parameter (42B active, 1M context) multimodal agentic coding model Muse Spark Open, InclusionAI's Ling 3.0 Flash, and OpenBMB's MiniCPM-V 4.7.

Google AI Studio Opens Gemini 4 Argon in Build & Playground for Pro Users; Transformers v5.20.1 Adds Native Reflection AI Beam and Meta 816B Muse Spark Open

Key Takeaways

  • Google AI Studio opens Gemini 4 Argon in AI Studio Build and Playground (https://ai.studio/build) for Google AI Pro subscribers
  • Hugging Face Transformers v5.20.1 natively integrates Reflection AI's token-efficient coding model Beam and fixes DeepGEMM Mega MoE under expert parallelism
  • Adds native support for Meta's 816B-parameter (42B active, 1M context, BF16/NVFP4) omni-modal coding model Muse Spark Open and InclusionAI's KDA+MLA hybrid Ling 3.0 Flash
Read full story
Linked Models & Agents:
#Gemini-4-Argon#Google-AI-Studio#HuggingFace-Transformers#Reflection-AI-Beam#Muse-Spark-Open#Ling-3.0-Flash
M
Meta Superintelligence Labs@MetaAI·
👀115
🔥97

Meta open-sources Muse Spark Open: 816B-total / 42B-active MoE with 1M context, billed as the most capable US open-weight model to date

On 2026-10-15 Meta released weights for Muse Spark Open, a natively multimodal MoE (~816B total / ~42B active per token, 1M context) aimed at long-horizon coding and agents. Weights are on Hugging Face and peers; hosted APIs include Together, Fireworks, Groq and OpenRouter. Reported figures include DeepSWE 60%, MCP-Atlas 89.2% and JobBench 61.8% (see Meta’s evaluation methodology).

Meta open-sources Muse Spark Open: 816B-total / 42B-active MoE with 1M context, billed as the most capable US open-weight model to date

Key Takeaways

  • Scale: ~816B total / ~42B active per token; 256 experts/layer with top-8 routing; 62 layers, hidden 8192; context up to 1,048,576 (HF blog)
  • Coding/agents (Meta scorecard / AA, xhigh): DeepSWE 60.0%; SWE-Atlas Codebase QnA 58.0%; MCP-Atlas 89.2%; JobBench 61.8%; Toolathlon-Verified 72.8% (eval methodology)
  • Throughput (NVIDIA tech blog, agentic on GB300 NVL72): up to ~50K tok/s/GPU and ~600 tok/s/user; NVFP4 checkpoints published (NVIDIA blog)
  • Deploy: Muse Model License with limited commercial terms; SGLang / vLLM / TensorRT-LLM / MLX / llama.cpp; Meta demos two Apple M5 Ultra Mac Studios for on-desk local run
  • Third-party index: Artificial Analysis Intelligence Index ~38.9–39 (xhigh, as of 2026-10-07)—behind closed frontier, but a key open-weight reference (AA)
Read full story
Linked Models & Agents:
⚡ Muse Spark Open⚡ Muse Spark Open 816B-A42B🤖 Muse Code🌐 Meta🌐 NVIDIA🌐 Hugging Face🌐 Apple
#Meta#Muse Spark Open#open weights#MoE#agentic coding#816B#MLX#NVFP4
A
AI at Meta@AIatMeta·
👀138
🔥98
View on 𝕏 ↗

Meta open-sources Muse Spark Open: 816B/42B natively multimodal US open-weight flagship, runs on two M5 Ultra Mac Studios

On 2026-10-15 Meta released Muse Spark Open open weights: an 816B-total / 42B-active MoE with a Perception Encoder, natively multimodal (text/image/video/audio in → text out), 1M context, aimed at coding and long-horizon agents. Weights are on Hugging Face (BF16/NVFP4), SageMaker, Model Garden and Foundry; APIs include OpenRouter, Together, Fireworks and Groq. Meta calls it the most capable US open-weight model to date and, with Apple, optimized it to run locally across two M5 Ultra Mac Studios via MLX.

Meta open-sources Muse Spark Open: 816B/42B natively multimodal US open-weight flagship, runs on two M5 Ultra Mac Studios

Key Takeaways

  • Specs: MoE 816B total / 42B active (256 experts, top-8), 62 layers, 1M context; text+image+video+audio in → text out (model card)
  • Local: NVFP4 needs ≥~700GB VRAM, BF16 ~2TB; or two Apple M5 Ultra Mac Studios via MLX (Meta blog)
  • Agentic benches (xhigh, official table): MCP-Atlas 89.2, JobBench 61.8, OSWorld Verified 80.4, Terminal-Bench 2.1 76.4, DeepSWE 1.1 60, AA Intelligence Index 39
  • Serve with SGLang / vLLM / TensorRT-LLM / Transformers / llama.cpp / Unsloth / MLX; Helion kernels claim 1.03–3.95× kernel speedups on NVIDIA/AMD
  • License: Muse Model License (display “Built with Muse Spark Open”; MaaS over $20M/yr needs a separate deal); OpenRouter ~$1.25/M in, $4.25/M out
Read full story
Linked Models & Agents:
⚡ Muse Spark Open⚡ Muse Spark🤖 muse-code🌐 Meta AI🌐 apple🌐 huggingface🌐 openrouter
#Meta#Muse Spark Open#MoE#Agents#Apple MLX#Mac Studio#1M context
N
NVIDIA AINVIDIAAI·
👀118
🔥96
View on 𝕏 ↗

NVIDIA open-sources Nemotron 3.5 Super VL (120B-A12B multimodal agent model, 1M context) and the 24GB 2-bit local Bolt 74B-A9B

On 2026-10-15 NVIDIA released Nemotron 3.5 Super VL, an open multimodal agent model with 120B total / 12B active parameters: a Mamba-2–Transformer hybrid MoE with a C-RADIO vision encoder, text/image/document/video input, up to 1M context, and built-in MTP plus LiLiCorr speculative decoding. Weights, datasets and recipes ship under OpenMDW-1.1 on Hugging Face, OpenRouter, ModelScope and NIM. Alongside it, NVIDIA published Bolt 74B-A9B 24GB, a natively 2-bit (QAT) local agent model for RTX/DGX Spark; the 4GB Bolt E4B follows on Oct 23.

NVIDIA open-sources Nemotron 3.5 Super VL (120B-A12B multimodal agent model, 1M context) and the 24GB 2-bit local Bolt 74B-A9B

Key Takeaways

  • Scale: Super VL is 120B total / 12B active with up to 1M context; BF16 reference serving needs 4×H100 or 2×GB200, with an NVFP4 build for smaller footprints (model card)
  • Coding (vendor-reported): SWE-bench Verified 71.10, Terminal-Bench 2.1 66.70, LiveCodeBench v6 88.22; in the same table Qwen3.5-122B-A10B scores 71.9 / 48.70 / 76.7 and GLM-5.3-Flash 79.10 / 82.70 / 80.6
  • Agent evals: Hermes Bench 63.44% (Nemotron 3 Ultra 56.33%), LangChain Evals 87.05% (Ultra 84.3%); NVIDIA claims up to 5x single-GPU throughput vs open peers and 600+ TPS on Baseten (tech blog)
  • Bolt 74B-A9B 24GB: native 2-bit (IQ2 + QAT), 256K context; Terminal-Bench 2.1 50.98 and BrowseComp 66.75 vs 44.38 / 48.74 for Qwen3.6-35B-A3B; PinchBench 87.82 is slightly below Qwen3.6's 88.07 (Bolt card)
  • Serving: LiLiCorr speculative decoding gives up to 2.3x faster generation on DGX Spark; the Dynamo recipe serves 8x more users with 85% lower TTFT on 16×B200 at 50 tok/s/user
Read full story
Linked Models & Agents:
⚡ nemotron-3.5-super-vl⚡ nemotron-3.5-bolt-74b-a9b⚡ nemotron-3.5-bolt-e4b⚡ Nemotron 3 Ultra🤖 hermes-agent🤖 opencode🤖 OpenHands (All-Hands AI)🤖 Cline (VS Code Agent)🌐 Nvidia (NIM)
#Nemotron 3.5#NVIDIA#MoE#Mamba#Agents#Hugging Face#OpenRouter
M
ModelScope@ModelScope2022·
👀139
🔥95
View on 𝕏 ↗

StepFun opens Step 5 Preview weights: 600B/27B sparse MoE agent flagship with 1M context now on Hugging Face and ModelScope

On 2026-10-15 StepFun released open weights for Step 5 Preview as promised: a 600B-total / 27B-active sparse MoE with up to 1M-token context and vision input, aimed at agentic software engineering and knowledge work. Checkpoints are on Hugging Face (FP8, BF16, NVFP4) and ModelScope. Official scores include StepCodeBench 49.0%, FrontierFinance 66.4%, and Artificial Analysis Intelligence Index 44. September coverage was API/index only; this is the first public weight drop.

StepFun opens Step 5 Preview weights: 600B/27B sparse MoE agent flagship with 1M context now on Hugging Face and ModelScope

Key Takeaways

  • Specs: sparse MoE 600B total / 27B active, up to 1M-token context, text+image input (model page)
  • Weights: stepfun-ai/Step-5-Preview plus BF16 and NVFP4 on Hugging Face and ModelScope (HF card)
  • Coding: StepCodeBench avg@4 49.0% (leads the open-weight set StepFun cites); DeepSWE v1.1 67.7% (official table)
  • Finance / index: FrontierFinance 66.4%; Artificial Analysis Intelligence Index 44 (matches earlier AA coverage)
  • Serve via official vLLM / SGLang Docker images and OpenAI-compatible APIs; license is StepFun Model License 1.0 (Apache-2.0-style with commercial conditions)
Read full story
Linked Models & Agents:
⚡ Step 5 Preview🌐 stepfun🌐 modelscope🌐 huggingface
#StepFun#Step 5 Preview#MoE#Agents#1M context#Hugging Face#ModelScope
ADSponsored
H
Hugging Face Daily Papers@HuggingFace·
👀255
🔥99

Xiaomi Unveils MiMo-V2.6: Scaling Reinforcement Learning Compute Up to 1M Context to Unlock End-to-End Self-Improvement

Reinforcement learning (RL) has emerged as the defining post-training paradigm driving frontier foundation models toward autonomous self-improvement, yet scaling RL across multi-step agent environments and million-token sequences introduces profound stability and infrastructure hurdles. Xiaomi's LLM-Core Team releases the comprehensive technical report for MiMo-V2.6 (arXiv:2610.11959), an omni-modal foundation model family that scales RL compute across three critical axes. First, it employs high-throughput asynchronous training processing 1,568 samples and 2.7 to 3.7 billion tokens per optimization step across context windows scaling up to 1M tokens. Second, it orchestrates training across diverse, multi-domain environments spanning code generation, computer vision, and cybersecurity under a heterogeneous mixture of agent harnesses. Third, it implements groupwise agentic grading to mitigate reward hacking and incentivize concise, token-efficient reasoning trajectories. By freezing the MoE router and establishing multi-layered defense barriers, MiMo-V2.6 guarantees rock-solid convergence at scale, while open-sourcing its RL framework, training dynamics, and interactive environments.

Key Takeaways

  • Xiaomi releases MiMo-V2.6, scaling reinforcement learning compute across 1,568 samples, 3.7B tokens per step, and up to 1M context
  • Introduces groupwise agentic grading to curb reward hacking and compress trajectory token lengths by over 35%
  • Freezes MoE routers to eliminate routing collapse and open-sources the complete RL post-training infrastructure
Read full story
Linked Models & Agents:
⚡ MiMo-V2.6⚡ Hybrid-SWA⚡ Agentic-RL⚡ MoE🤖 Xiaomi-LLM-Core🤖 Groupwise-Grader🤖 Self-Improving-Agent🌐 Xiaomi🌐 Hugging Face
#MiMo-V2.6#Xiaomi#Reinforcement Learning#Self-Improvement#1M Context#Open Source
M
Mistral AI@MistralAI·
👀396
🔥96

Mistral Large 4 "Le Chonk" public preview: 1.05T-total / 49B-active multimodal MoE, 1M context, open weights by month-end

On 2026-10-06 Mistral launched a public-preview API for Mistral Large 4: a granular MoE with 1.05T total / 49B active parameters plus a 1.6B vision encoder, 1M context, trained from scratch on 3,800 Grace Blackwell GPUs in Mistral's own European datacenters. It posts 61.7% DeepSWE v1.1 and 93% Cybench; weights are promised by the end of October.

Mistral Large 4 "Le Chonk" public preview: 1.05T-total / 49B-active multimodal MoE, 1M context, open weights by month-end

Key Takeaways

  • Scale: 1.05T total / 49B active MoE + 1.6B vision encoder, 1M context
  • Coding (Artificial Analysis private eval): DeepSWE v1.1 61.7%, SWE-Atlas-QnA 59.4%, Terminal-Bench 4 28.3%, Coding Agent Index 49.8%
  • Security: Cybench 93%; 82% on AA Cyber Index reproduce-and-patch test (highest of any model); Lakera B3 93.3% attack resistance
  • Agents: AutomationBench 59.9%, AA-Briefcase 1,393 Elo; Surge AI blind coding eval 3.74, behind only Claude Opus 5 (4.22)
  • Pricing per 1M tokens: $1.36 input / $0.14 cached / $4.18 output (docs also list a half-price tier); weights by end of month
Read full story
Linked Models & Agents:
⚡ Mistral Large 4🌐 Mistral AI
#Mistral#Mistral Large 4#Open weight#MoE#Multimodal#Cybersecurity#Foundation model
H
Hugging Face / Reflectionhuggingface.co·
👀310
🔥98

Reflection Beam weights are live: 501B/23B MoE under Apache 2.0, BF16/FP8/NVFP4 on Hugging Face, 80.1% Terminal-Bench 2.1

On 2026-10-16 Hugging Face and Reflection confirmed Beam weights are downloadable: ~501B-total / ~23B-active sparse MoE, 1M context, Apache 2.0. BF16/FP8/NVFP4 checkpoints ship with day-0 transformers≥5.20.1, vLLM and SGLang. Model-card figures (vendor): SWE-bench Verified 81.6% (BF16), Terminal-Bench 2.1 80.1%, MCP Atlas 80.8%. This closes the Oct 5 promise to release weights later that month.

Reflection Beam weights are live: 501B/23B MoE under Apache 2.0, BF16/FP8/NVFP4 on Hugging Face, 80.1% Terminal-Bench 2.1

Key Takeaways

  • Weights live: reflection/Beam-1.0-501B-A23B plus FP8/NVFP4; Apache 2.0 (HF blog)
  • Scale: ~501B total / 23B active; 52 layers (3 dense + 49 MoE), 256 experts, top-8; hybrid sliding-window + global attention; 1M context
  • Coding/terminal (model card BF16, temp=0.7): SWE-bench Verified 81.6%; Terminal-Bench 2.1 80.1%; DeepSWE v1.1 43.3%; MCP Atlas 80.8% (vendor-reported)
  • Efficiency: Reflection claims ~3× fewer inference reasoning tokens vs leading open peers; effort low–max with preserve_thinking
  • Deploy: vLLM/SGLang recipes with DFlash2; BF16 ~1TB, NVFP4 ~280GB; plan ≥4×B200; try HF Inference Providers (Baseten/Fireworks/Together)
Read full story
Linked Models & Agents:
#Reflection#Beam#open weights#Apache 2.0#MoE#501B#Hugging Face#agentic coding#weights live
C
Cursor@cursor_ai·
👀95
🔥92

Cursor lists GLM 5.3 / Flash natively: 1M context, Other Models pool, from $1.4 / $0.15 per MTok

Cursor’s official Models & Pricing and dedicated docs now list Z.ai GLM 5.3 and GLM 5.3 Flash as addable third-party models: 1M context, Low/High/Max reasoning (High default), Other Models pool billing. Table rates: GLM 5.3 $1.4/$0.26/$4.4, Flash $0.15/$0.029/$0.5 per MTok—same as 5.2, no long-context surcharge. Enable in Settings → Models (not the BYOK custom-endpoint path).

Key Takeaways

Read full story
Linked Models & Agents:
⚡ GLM-5.3⚡ GLM-5.3-Flash⚡ GLM-5.2🤖 Cursor Composer (Agent Mode)🌐 Cursor (Anysphere)🌐 zhipu🌐 z-ai
#Cursor#GLM 5.3#GLM 5.3 Flash#Z.ai#Zhipu#model picker#Other Models#1M context#Open weight
A
Anthropic@AnthropicAI·
👀305
🔥94

Claude Code 2.1.287: Mods in-process extensions; built-in You should know side agent; 1M context default on gateways

Anthropic shipped Claude Code 2.1.287 on 2026-10-01 (npm @anthropic-ai/[email protected]): Claude Mods let plugins run in-process JS/TS hooks that rewrite tool calls and UI; opt-in built-in You should know side agent; MCP URL elicitation; Opus 4.7+/Fable default to 1M context on Bedrock/Vertex/Foundry/apps gateway. Agent SDK TS 0.3.287 parity.

Claude Code 2.1.287: Mods in-process extensions; built-in You should know side agent; 1M context default on gateways

Key Takeaways

  • Release v2.1.287; npm 2.1.287; GitHub ~2026-10-01T18:00:22Z
  • Mods: in-process hooks for UI + tool.call rewrite; ≥2.1.287; docs https://code.claude.com/docs/en/plugins/mods/overview
  • You should know: /plugin enable cc-plugin-you-should-know@builtin (first-party + telemetry); off by default
  • 1M context default for Opus 4.7+/Fable on Bedrock/Vertex/Foundry/apps gateway; disable via CLAUDE_CODE_DISABLE_1M_CONTEXT=1
  • Agent SDK TS 0.3.287; MCP URL elicitation / bareElicitationCapability
Read full story
Linked Models & Agents:
⚡ Claude Opus 5.5⚡ Claude Sonnet 5.5⚡ Claude Fable 5.1🤖 Claude Code (Anthropic)🌐 Anthropic
#Claude Code#Mods#You should know#Plugins#MCP elicitation#1M context#Agent SDK#2.1.287