Issue 15 · Week 31 of 2026.

Inkling-Small joins the tree as the strongest US open-weights efficiency result to date, reaching an independent Intelligence Index of 40 against Inkling's 41 with under a third of the parameters. Kimi K3 is reclassified from closed to open weights on the July 27 artifacts and license text, with its active-parameter count corrected from a community estimate to a confirmed 104B. DeepSeek V4 Flash updates for the 0731 re-post-train checkpoint that gained 10 independent Index points on unchanged architecture.

Kimi K3 is an update rather than an addition: the row has existed since W29, and this week's material change is openness, license terms, and a corrected active-parameter count of 104B against the earlier community-estimated ~50B. Qwen3.7 Flash is held out of the tree pending a first-party model card, and MiniMax H3 is deferred until its promised weights actually ship. Both are covered in the cards.

Vendor-stated frontier capability. The releases that reset the closed-source ceiling.

• 2026-07-30/OpenAI/Frontier/Reasoning

  GPT-5.6 Luna / Terra price cut and Sol Fast mode

  Luna falls 80% to $0.20/$1.20, Terra 20% to $2/$12, Sol holds $5/$30, and Fast mode sells ~2.5x speed at 2x token price

  Re-baseline high-volume routing immediately: Luna undercuts most Western cheap tiers on sticker price while Sol remains the premium band against Opus 5. Treat Fast mode as a purchasable latency tier rather than a separate model, and instrument per-tier completion rates before moving fleet volume on list price alone. The AI Stack Weekly owns the fixed-workload spread calculation; this Pulse stays on the architecture and routing implication.

  OpenAI; Unite.AI; VentureBeat; The Decoder

• 2026-07-31/DeepSeek/Frontier/MoE

  DeepSeek V4-Flash-0731

  Same 284B/13B MoE and $0.14/$0.28 pricing, re-post-trained into public beta; independent AA Index jumps 10 points to 50

  Drop this into existing Flash routes as a cheap-tier agent upgrade without a size or price change — but do not treat it as V4-Pro GA, and pin the harness when comparing Terminal-Bench (vendor-claimed 82.7 max effort versus Artificial Analysis 79%). Architecture and model ID stay deepseek-v4-flash; Pro, app, and web remain unchanged pending the promised official V4-Pro.

  DeepSeek API changelog; Artificial Analysis (independent)

• 2026-07-29/OpenAI/Reasoning/Agentic

  GPT-5.6 Sol harness memory on ARC-AGI-3

  Vendor-reported public-set score triples from 13.3% to 38.3% via retained reasoning plus context compaction on the same model

  Treat the ARC-AGI-3 jump as vendor-claimed on a single harness, not an official leaderboard reprint: OpenAI reports roughly 6x fewer output tokens when compaction replaces rolling truncation on the Responses API, and no qualifying LMArena or SWE-bench official moves landed in-window. Agent Techniques Weekly owns the method and the reproduction caveat.

  OpenAI (vendor-reported harness result)

• 2026-07-27/Alibaba / Qwen/Specialist/Multimodal

  Qwen3.7 Flash

  Quiet low-cost multimodal API at reported $0.03/$0.13 under 32K and 1M context — no lab blog, no weights, no public benchmark table

  Useful as a fleet-cost floor candidate for multimodal agents, but treat specs and pricing as secondary until a first-party model card is pinned. Sources are OpenRouter listing dates and technical digests citing a QwenCloud changelog; no qwenlm.github.io launch post was found in-window.

  eesel.ai; BenchLM (secondary; no primary lab announcement)

Open-weights releases that change procurement options. Pull these into pilot when score parity meets license parity.

• 2026-07-27/Moonshot AI/Open frontier/MoE

  Kimi K3

  Full weights ship: 2.8T MoE, 104B active, 896/16 experts, hybrid 69 KDA + 24 Gated MLA, 1,048,576 context, revenue-tiered license

  Start the self-host evaluation now — GitHub and Hugging Face artifacts landed Jul 27 — but send the Kimi K3 License to counsel before any commercial hosting plan: MaaS above $20M aggregate trailing-12-month revenue needs a separate deal, with attribution triggers above 100M MAU or $20M monthly revenue. Independent AA Index reference remains 57; vendor-claimed table includes GPQA Diamond 93.5 and DeepSWE 67.5. Hosted API stays $3/$15 (cached input $0.30).

  MoonshotAI/Kimi-K3 GitHub; Hugging Face; license text; Unite.AI

• 2026-07-30/Thinking Machines Lab/Open frontier/MoE

  Inkling-Small

  276B/12B-active multimodal MoE under Apache 2.0; independent AA Index 40 vs Inkling 41 at under a third the parameters

  Strongest US open-weights efficiency play this week for teams that cannot absorb K3-class cluster scale: BF16 needs ~600 GB aggregate VRAM, NVFP4 ~180 GB, with vLLM v0.26.0 shipping same-week Inkling family support. Flag the context conflict honestly — vendor model card says up to 1M; Artificial Analysis lists 256K in its article blurb — and use Index 40 as the independent baseline rather than vendor-claimed SWE-bench Verified 80.2% alone.

  Thinking Machines Lab; Artificial Analysis (independent Index 40); VentureBeat

• 2026-07-31/MiniMax/Specialist/Multimodal

  MiniMax H3

  Multimodal video model launches API-first with open weights promised in coming days — still not on Hugging Face or GitHub as of Aug 1

  Treat current state as API-only until the repo appears: unified text/image/video/audio understanding to video with native stereo, up to 15s at 2K, with vendor claims that 2K per-second price is under one-third of mainstream models. Do not build a self-host plan around the promise; the W30 K3 lesson applies again.

  MiniMax blog; SCMP; secondary weights-not-yet status

Architectural patterns that crossed multiple vendors this period. Each pattern lists exemplar releases and what it changes for deployment, cost, or capability.

• Post-training and harness policy move scores without new weights

  DeepSeek V4-Flash-0731: +10 AA Index on unchanged 284B/13BOpenAI ARC-AGI-3: 13.3% to 38.3% via retained reasoning + compactionOpenAI Jul 29 serving/harness efficiency post (~20% serving cost, >15% token-gen efficiency)

  Two independent vendors moved measured capability this week without shipping new parameter counts: DeepSeek re-post-trained Flash in place, and OpenAI changed Responses API memory settings on the same Sol weights. Architects should treat reasoning retention, context compaction, and post-training checkpoints as first-class evaluation dimensions alongside the model ID. Buyers who only diff weight releases will miss the largest week-over-week gains.

  DeepSeek; Artificial Analysis; OpenAI

• Revenue-tiered open licenses turn legal review into the deployment gate

  Kimi K3 License: >$20M trailing-12-month MaaS commercial triggerKimi K3 attribution triggers above 100M MAU or $20M monthly revenueInkling-Small: clean Apache 2.0 contrast case

  Downloadable weights no longer imply a single open-source diligence checklist. K3 is MIT-like until commercial hosting crosses the revenue trigger, at which point counsel must negotiate; Inkling-Small remains Apache 2.0 with no such trigger. Procurement should classify each open release by license economics before sizing clusters, and keep Apache-clean alternatives on the shortlist when MaaS revenue is material.

  MoonshotAI/Kimi-K3 LICENSE; Thinking Machines Lab

• Serving stacks ship in the same week as the open MoE class

  vLLM v0.26.0: Inkling modeling stack, NVFP4, MTP speculative decodingInkling-Small BF16 ~600 GB / NVFP4 ~180 GB VRAM floorsDeepSeek-V4 routing and speculative optimizations across NVIDIA/AMD/XPU

  Production self-hosters got a same-week inference path for the new open MoE class rather than a multi-month serving lag. vLLM v0.26.0 (Jul 27) adds Inkling CUDA-graph and Hopper FA4 relative attention support alongside DeepSeek-V4 serving pushes and matured KV offloading. Capacity planners should price FP4 datapaths and KV-tiered storage into the model-release decision, not as a later optimization.

  vLLM project GitHub release v0.26.0; Thinking Machines model card

• Near-iso Intelligence at a fraction of active parameters

  Inkling-Small: AA Index 40 vs Inkling 41 at <1/3 parametersInkling-Small: 12B active vs Inkling 41B activeDeepSeek V4 Flash prior: also sat at AA Index 40 in the same size class

  Artificial Analysis places Inkling-Small within one Index point of full Inkling while activating 12B rather than 41B parameters, and notes no smaller open-weights model scores higher in that size class. Efficiency MoEs are now a procurement tier of their own: teams should ask whether the last Index point is worth 3x the serving footprint before defaulting to the flagship open checkpoint.

  Artificial Analysis (independent)

Benchmark deltas that change a procurement read. Scores reflect public leaderboards or vendor model cards as of publication.

• Artificial Analysis Intelligence Index — DeepSeek V4 Flash

  Independent Index jumps from 40 to 50 (+10) for Flash-0731 on unchanged architecture; now ~1 point behind GPT-5.6 Luna (51) and 6 points above V4 Pro on AA's comparison

  • DeepSeek V4 Flash 0731 (max)50 Index; GDPval-AA v2 1559 Elo; Terminal-Bench 2.1 79% (AA)
  • Prior DeepSeek V4 Flash40 Index; GDPval-AA v2 1189 Elo
  • Kimi K3 (max)57 — open-frontier reference
  • GPT-5.6 Luna (max)51
  • GLM-5.2 (max)51

  Artificial Analysis (independent)

• Artificial Analysis Intelligence Index — Inkling-Small

  New open US model lands at Index 40, within 1 point of Inkling (41) at less than a third the parameters

  • Inkling-Small40 (independent)
  • Inkling41 (independent)
  • DeepSeek V4 Flash (prior max, same size class)40
  • MiniMax-M2.7 (per AA)~38 — about 2 points below Inkling-Small

  Artificial Analysis (independent)

• ARC-AGI-3 public set (vendor harness, not official board)

  OpenAI reports GPT-5.6 Sol rising from 13.3% to 38.3% by retaining reasoning and compacting context — same model, roughly 6x fewer output tokens

  • GPT-5.6 Sol + retained reasoning + compaction38.3% (vendor-reported)
  • GPT-5.6 Sol + rolling truncation baseline13.3% (vendor-reported)

  OpenAI (vendor-reported; not an official ARC leaderboard reprint)

• DeepSeek Terminal-Bench 2.1 (vendor harness vs independent)

  Vendor-claimed max-effort 82.7 versus Artificial Analysis 79% on the same Flash-0731 checkpoint — pin the harness before ranking

  • DeepSeek V4 Flash 0731 (DeepSeek harness, max effort)Terminal-Bench 2.1 82.7 (vendor-claimed)
  • DeepSeek V4 Flash 0731 (Artificial Analysis)Terminal-Bench 2.1 79% (independent)
  • DeepSeek V4 Flash 0731 (DeepSeek harness)Cybergym 76.7; Toolathlon verified 70.3; DeepSWE 54.4 (vendor-claimed)

  DeepSeek API changelog (vendor); Artificial Analysis (independent)

A snapshot of leader-vs-challenger by tier. Useful for procurement shortlists when matching workload to model class. Pair with the benchmark moves above for the underlying scores.

• Closed frontier

  Leader: Claude Fable 5

  Challenger: Claude Opus 5

  No new closed flagship shipped; Fable retains the absolute reference while Opus 5 remains the economic default at $5/$25.

• Open frontier

  Leader: Kimi K3

  Challenger: Inkling-Small

  K3 takes the lead now that weights and license text are public (AA Index 57); Inkling-Small is the efficiency challenger at Index 40 under Apache 2.0.

• Reasoning

  Leader: Claude Fable 5

  Challenger: GPT-5.6 Sol

  Sol's vendor-reported ARC-AGI-3 harness jump shows memory policy can move scores without new weights; independent cost-per-task still pending.

• Coding

  Leader: Claude Opus 5

  Challenger: DeepSeek V4 Flash 0731

  Opus 5 holds the practical coding lead on W30 launch evidence; Flash-0731 is the cheap-tier agent upgrade at Index 50 and unchanged $0.14/$0.28.

• Multimodal

  Leader: Gemini 3.6 Flash

  Challenger: Kimi K3

  Gemini remains the public fleet multimodal default; K3 now adds self-host control with MoonViT-V2. Gemini 3.5 Pro still has not GA'd.

• Edge / small

  Leader: Inkling-Small

  Challenger: Gemini 3.5 Flash-Lite

  Inkling-Small brings open multimodal agents to ~180 GB NVFP4; Flash-Lite remains the hosted speed-cost floor at roughly 350 tok/s.

The non-release moves that shift vendor risk — pricing, deprecations, gating decisions, license changes — with a one-line procurement read.

• 2026-07-30/OpenAI

  GPT-5.6 Luna cut 80% to $0.20/$1.20 and Terra 20% to $2/$12; Sol held at $5/$30; Fast mode replaces Priority Processing

  Closed labs are competing on completed-task economics via rate cards and latency tiers rather than new flagship weights this week. Route only latency-sensitive work into Fast mode and keep ordinary asynchronous agents on standard Luna or Terra after measuring completion quality.

  OpenAI

• 2026-07-27/Moonshot AI

  Kimi K3 License ships with weights: MIT-like until MaaS revenue exceeds $20M trailing twelve months

  Open-weights diligence now splits into quality evaluation and license classification. Counsel should map expected MaaS and product revenue against the trigger before any production self-host commitment; Apache-clean alternatives remain available for teams that cannot accept the commercial clause.

  MoonshotAI/Kimi-K3 LICENSE; Hugging Face

• 2026-07-31/Google DeepMind

  Gemini 3.5 Pro still not generally available as the standing Jul 31 prediction deadline expires

  Do not plan Q3 procurement around 3.5 Pro. Buy the callable Flash and Flash-Lite fleet now; another slip further weakens roadmap-based purchasing.

  Google (Jul 21 partner-testing status; no in-window GA found)

• 2026-07-31/MiniMax

  H3 launches with an explicit promise to open weights in the coming days; no public HF or GitHub artifacts as of Aug 1

  Repeat the K3 lesson: an open-weights commitment is not a downloadable artifact. Keep H3 on the API watchlist until the repo and license appear.

  MiniMax blog; secondary weights-not-yet reports

• 2026-07-27/vLLM project

  vLLM v0.26.0 ships Inkling family support, NVFP4, MTP speculative decoding, and DeepSeek-V4 serving optimizations

  Serving readiness is now part of the model-release race. Self-host teams evaluating Inkling-Small or DeepSeek V4 this week should pin v0.26.0 as the baseline inference path rather than waiting for a later stack catch-up.

  vLLM GitHub releases

Specific model-side catalysts in the next 7–30 days that would change the read materially. Watching these tells us whether the canopy is widening or thinning.

• Next 7-14 days

  MiniMax H3 downloadable weights and license

  Decides whether H3 becomes a real open video-generation option or remains an API product wrapped in open-language marketing.

• Next 30 days

  DeepSeek V4-Pro official GA and pricing

  Flash-0731 is not Pro GA; the standing surge-pricing and official Pro triggers remain incomplete until a callable Pro ID and rate card land.

• Next 14 days

  Inkling-Small context window: vendor 1M vs AA 256K

  Procurement specs need a resolved context number; watch for an AA methodology note or Thinking Machines clarification.

• By Aug 15

  Independent Opus 5 cost-per-task results

  Vendor benchmark leadership from W30 is still not enough; the market needs task cost, token use, and latency against Fable 5 and GPT-5.6 Sol.

• Next 2-3 weeks

  LMArena Opus 5 Elo stabilization

  No qualifying in-window official board move; secondary pages remain unstable — wait for a date-stamped stable print before re-ranking coding.

• Next 30 days

  Gemini 3.5 Pro callable model ID and public pricing

  The Jul 31 miss is now a track-record item; only a public GA closes the W30 prediction cleanly.