Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 

README.md

Eight Specialized AI Model Types: Landscape & ZAO Fleet Relevance

Doc 1185 | Research | Updated 2026-07-16 | Type: Reference | Tier: STANDARD


Overview

Zaal shared an infographic on "8 Different Specialized AI Models" (LLM, LCM, LAM, MoE, VLM, SLM, MLM, SAM) and asked if we've researched them. This doc surveys all 8 with 2026 current examples, use cases, and concrete analysis of which matter to the ZAO agent fleet. The landscape shifted dramatically in 2025-2026: MoE became standard in frontier models, SLMs proved viable for edge, and LAMs (agent-first) emerged as a distinct category. This research grounds the fleet's model routing strategy (ZOL Model Gateway, ZOE's worker dispatch, Ollama local inference).


The 8 Model Types: What, How, Examples, Best At, Limits

1. LLM - Large Language Model

What it is: Autoregressive transformers (next-token prediction) trained on massive text corpora to generate coherent text. The foundation of the AI wave (2017-2026).

Pipeline:

  1. Tokenization (BPE/SentencePiece) converts text to subword IDs
  2. Embedding layer maps token IDs to dense vectors
  3. Transformer stack (multi-head attention, feed-forward) processes sequence bidirectionally (encoding only in LLM)
  4. Output projection predicts next token probability distribution
  5. Sampling/beam-search decodes into natural language

2026 Examples:

  • OpenAI GPT-4o, GPT-5 (frontier, closed)
  • Anthropic Claude Opus 4.7 (frontier, conversation/reasoning)
  • Meta Llama 3.2 (70B open-weight, strong baseline)
  • Mistral Large 2 (open-weight, cost-effective)

Best at:

  • Long-form generation (email, code, prose, reasoning)
  • Multi-turn conversation (in-context learning)
  • Zero-shot transfer (handles unseen tasks)
  • Instruction following (fine-tuning via RLHF or synthetic data)

Limits:

  • Context window (even 200K tokens hit cost + latency cliffs)
  • Hallucination (confident false outputs, especially on retrieval)
  • Slow inference (token-by-token autoregressive)
  • Cost at scale (frontier models $5-50/MTok input)
  • No structured reasoning (reasoning is implicit in weights)

2. LCM - Large Concept Model

What it is: Meta's paradigm shift: operate on explicit sentence-level semantic embeddings (SONAR space) instead of tokens. Predicts entire sentences in a language- and modality-agnostic concept space.

Pipeline:

  1. Input text → SONAR encoder (200+ languages, dual text+speech)
  2. Concept space: each token → sentence embedding (e.g., 1024-dim semantic vector)
  3. LCM transformer (1.6B-7B params) predicts next concept given prior concepts
  4. Decoder maps predicted concept back to text (any language)

2026 Examples:

  • Meta LCM (7B, outperforms Llama 7B on multilingual tasks)
  • v-LCM (vision-language, extends SONAR to vision via v-Sonar)
  • v-Sonar embedding space (1500 text languages + 177 speech languages, unified)

Best at:

  • Multilingual generation (concept space is language-agnostic)
  • Cross-lingual transfer (train on one language, zero-shot generalize to 61 others)
  • Long-document reasoning (sentences as units, not tokens, = fewer steps)
  • Semantic search (SONAR embeddings are highly interpretable)

Limits:

  • Immature ecosystem (no fine-tuning harness, no production deployments yet)
  • Concept → text decoder bottleneck (reversing embeddings loses information)
  • Adoption slow (industry still token-first)
  • Emerging (v-LCM only released 2025, still research-grade)
  • Not widely available via API (Meta research only)

3. LAM - Large Action Model

What it is: Agent-first architecture: perception → intent recognition → task decomposition → action planning → tool execution. Agents that do things, not just predict text. Goes beyond "output a function call" (LLM tool-use) to stateful, multi-step task ownership.

Pipeline:

  1. Perception: parse user input + screen state + prior context
  2. Intent recognition: classify whether task requires action or just text
  3. Decompose: break complex goal into substeps
  4. Plan: sequence actions (click, type, API call, navigate)
  5. Execute: run actions in environment, observe feedback
  6. Learn: update state, adjust next action based on result
  7. Repeat until task done or escalate to human

2026 Examples:

  • Rabbit R1's DLAM (transforms r1 into Windows/Mac computer controller)
  • Rabbit OpenClaw (open-source agent, runs in user's environment, direct action)
  • AI21's agent stack (perception-to-action middleware)

Best at:

  • Multi-step automation (book flights, file expenses, manage spreadsheets)
  • Computer vision tasks + action (screenshot → understand → click/type)
  • Embodied reasoning (agents that learn by doing)
  • Tool chaining (coordinate multiple APIs/software)

Limits:

  • Hallucination + action = real risk (agent books flight to wrong city)
  • Latency (perception → decomposition → execution is slow)
  • Brittleness (environment changes break trained behaviors)
  • Safety (agents need hard guardrails, not just prompts; see rule 9: one instance per resource)
  • Still in 2026: Rabbit's own LAM still makes mistakes, rates ~60-70% task accuracy on real workflows

4. MoE - Mixture of Experts

What it is: Sparse activation architecture: instead of running all parameters for every token, split network into K experts + router. Router selects top-N experts per token; only those N experts run. Stays fast while scaling parameter count.

Pipeline:

  1. Input token → gating network (router)
  2. Router outputs probability for each expert (linear + softmax, or cosine similarity)
  3. Top-K selection (e.g., top-2 for Mixtral, top-8 for DeepSeek-V3)
  4. Selected experts process token in parallel
  5. Weighted combination (router weights + expert outputs) → output
  6. Load balancing: auxiliary loss (or DeepSeek's dynamic bias) keeps expert utilization even

2026 Examples:

  • Mixtral 8x7B (top-2 across 8 experts, runs at 13B speed)
  • DeepSeek-V3 (top-8 across 256 tiny experts, 671B params, state-of-the-art reasoning)
  • GPT-4 (confirmed sparse MoE, exact architecture proprietary)
  • Qwen3-MoE (top-6 across 60+ experts)

Best at:

  • Cost-efficient scaling (650B params, inference cost of 100B model)
  • Task specialization (experts learn different domains)
  • Reasoning + code (DeepSeek-V3 dominates benchmarks)
  • Heterogeneous workloads (route coding tasks to "code expert," reasoning to "reasoning expert")

Limits:

  • Load balancing still unsolved (uneven token routing → GPU idle time, 2026 research active)
  • Communication overhead (inter-GPU expert sync)
  • Training instability (convergence slower than dense LLMs)
  • Inference optimization hard (not all hardware can parallelize expert selection)
  • Expert dead-outs (some experts unused in practice, wasted params)

5. VLM - Vision Language Model

What it is: Dual encoders (vision + text) + projection layer. Learns joint embedding of images and text, enabling image understanding + captioning + visual QA.

Pipeline:

  1. Image → vision transformer (ViT, e.g., CLIP's image encoder)
  2. Text → text transformer (BERT-style or LLM encoder)
  3. Projection: both outputs → shared embedding space
  4. Contrastive training: maximize similarity for (image, description) pairs, minimize for mismatches
  5. At inference: encode image + query text, rank by similarity

2026 Examples:

  • GPT-4o / GPT-5o (unified text+image, strong OCR + reasoning)
  • Claude Opus 4.7 (strong on document QA, computer-use GUI control)
  • Qwen2.5-VL (open-weight, competitive with GPT-4o on MMMU)
  • Pixtral (Mistral's VLM, fast, open-weight)
  • Llama 4 multimodal (open, strong on charts/diagrams)

Best at:

  • Visual reasoning (math, diagrams, flowcharts)
  • Document processing (PDF QA, receipt parsing)
  • Computer vision for agents (screenshot → understand → action)
  • Zero-shot visual search (no fine-tuning needed)
  • OCR + layout understanding (stronger than pure CV models)

Limits:

  • Image size limits (most cap at ~1-2MB encoded, low detail for very large images)
  • Latency (encoding + attention over image patches = expensive)
  • Token budget (images take 300-1500 tokens vs text's dense meaning)
  • Still hallucinate (confident wrong captions for ambiguous images)
  • Not specialized for real computer-vision tasks (edge detection, instance segmentation—SAM is better)

6. SLM - Small Language Model

What it is: Compact models (1B-13B params) designed for on-device or edge inference. Trade frontier performance for latency + cost + privacy.

Pipeline: Same as LLM (transformer, autoregressive), but:

  1. Smaller embedding dimension (512-1024 vs 8192 for frontier)
  2. Fewer layers (12-24 vs 80+ for frontier)
  3. Lower vocab size (sometimes)
  4. Quantization to 4-8 bit for memory efficiency

2026 Examples:

  • Phi-4 (14B, synthetic data magic, rivals 70B models on reasoning)
  • Gemma 4 (12B multimodal, fits ~7GB at 4-bit, text+image+audio+video)
  • Llama 3.2 (1B/3B, mobile-targeted, can run on smartphone)
  • Qwen-small (2.7B, competitive on code/math for size)

Best at:

  • Edge/on-device (no cloud latency, no privacy leak)
  • Cost efficiency (5-20x cheaper than LLM APIs for high volume)
  • Low-latency inference (2-5 tokens/sec on CPU)
  • Constrained devices (Raspberry Pi, iOS phone, embedded systems)
  • Specialized tasks with fine-tuning (after pre-train, customize for domain)

Limits:

  • Lower quality (Phi-4 is good, but not Claude Opus level)
  • Smaller context (often 4K-8K vs 200K+ for frontier)
  • Limited reasoning (few-shot weak, need examples in prompt)
  • Hallucination worse (smaller model, less parameter diversity)
  • Fine-tuning instability (small models overfit quickly on small datasets)

7. MLM - Masked Language Model

What it is: Bidirectional encoder trained by masking random tokens and predicting them from context. Powers retrieval, similarity, and reranking (not generation).

Pipeline:

  1. Input text → tokenize
  2. Randomly mask ~15% of tokens (replace with [MASK])
  3. Encoder processes full sequence (bidirectional attention)
  4. Output: predict the masked tokens' identities
  5. At inference: encode without masking, use hidden states as semantic vectors

2026 Examples:

  • BERT (10B, still used for retrieval)
  • BGE-M3 (embedding model, MIT-licensed, 100+ languages, sparse+dense+multi-vector)
  • E5 family (open-source, strong on semantic search)
  • SBERT / Sentence-BERT (encodes sentences into meaningful vectors)

Best at:

  • Semantic search (query → embedding, document → embedding, rank by similarity)
  • Clustering (group similar documents)
  • Similarity scoring (reranking LLM candidates)
  • Retrieval-augmented generation (encode corpus, find relevant docs fast)
  • Multilingual transfer (one encoder, cross-lingual search)

Limits:

  • No generation (can't write, only understand)
  • Frozen vocabulary (can't handle out-of-vocab words well)
  • Smaller models (encoder-only, so 110M-335M params typical vs 7B+ for SLM)
  • Sentence length (most cap at 512 tokens, long docs need chunking)
  • Training expensive (need large clean corpus, careful masking strategy)

8. SAM - Segment Anything Model

What it is: Prompt-driven image segmentation. Given an image + prompt (point click, bounding box, text, or exemplar image), output precise pixel-level mask of the object.

Pipeline:

  1. Image → vision encoder (ViT-based)
  2. Prompt (point/box/mask/text) → prompt encoder
  3. Mask decoder (lightweight, CNN-style)
  4. Decoder fuses image + prompt embeddings → probability map for each pixel
  5. Output: binary segmentation mask + confidence score

2026 Examples:

  • SAM (original, 2023)
  • SAM 2 (unified image+video, first-click refinement)
  • SAM 3 (concept-based: text prompts like "yellow school bus" segment all instances)
  • SAM 3.1 (real-time video, multiplexed tracking)

Best at:

  • Interactive segmentation (user clicks + refines)
  • Zero-shot segmentation (unseen object categories work)
  • Video object tracking (SAM 2+)
  • Data labeling automation (segment unlabeled images for annotation)
  • Concept discovery (SAM 3: "find all instances of X" without labeling)

Limits:

  • Not for semantic segmentation (which category? requires post-processing)
  • Small object weakness (SAM struggles with fine details < 10 pixels)
  • Prompt dependency (quality varies by prompt quality)
  • Compute (inference slow on edge, better on GPU)
  • Not for dense scene understanding (street scene with 50 objects needs 50 prompts)

ZAO Fleet Context: Current State

From CLAUDE.md + memory (ZOL, ZOE, Ollama):

Component Model Type Status Role
ZOE orchestrator LLM (Sonnet/Haiku) Live, production Concierge, task decomposition, worker dispatch
ZOL (@zolbot) LLM (Sonnet/Haiku via OpenRouter) Live, autonomous posts Farcaster agent, music scout persona
Ollama on VPS SLM (llama3.1:8b) Live, classification Inbox labeling, entity first-pass, health checks
Bonfire memory MLM (embeddings) Dormant, RLS block Knowledge graph recall, multi-corpus ingest
ZOL Model Gateway (mentioned in doc 1098, not yet read) MoE routing Planned Route tasks to right model based on cost/capability
Computer-use on VPS VLM (Claude vision) Planned, research Screenshot understanding for agents
Pi edge inference SLM + MoE Research Fallback when cloud unreachable

Comparison Table: Model Type Cheat Sheet

Model Type Pipeline Unit Params (typical) Inference Cost Best For Worst For
LLM Token 7B-670B $1-50/MTok Generation, reasoning, chat Fast retrieval, embedded devices
LCM Sentence 1.6B-7B $0.20/MTok (est.) Multilingual, cross-lingual zero-shot Production maturity, fine-tuning
LAM Action N/A (composition) $2-10/task (est.) Computer automation, multi-step tasks Single-turn queries, hallucination risk
MoE Token (sparse) 40B-670B total $0.50-5/MTok Reasoning + cost efficiency, scaling Load balancing, inter-GPU sync
VLM Image+text 7B-130B $2-20/image Visual reasoning, document QA, GUI control Fine-grained segmentation, edge deployment
SLM Token 1B-13B $0.05-0.50/MTok On-device, edge, cost (5-20x cheaper) Frontier reasoning, large context
MLM Sentence 110M-335M $0.001-0.01/embedding Semantic search, clustering, reranking Generation, long documents (>512 tokens)
SAM Pixel mask 600M-2B $0.10-0.50/image Interactive segmentation, zero-shot masking Semantic segmentation, dense scenes

Cross-Reference: Which 8 Types Matter to ZAO Fleet?

Verdict Matrix

Type Current Use ZAO Relevance Verdict Reason
LLM ZOE, ZOL core Critical USE NOW Foundation: all dispatch, reasoning, concierge run on Claude/Haiku LLMs. No alternative.
LCM None yet Speculative MONITOR Multilingual community bridges (Apna, Crypto Factor AVAX, Berlin partners) may need cross-lingual zero-shot. v-LCM maturity = 2026, tooling immature. Watch for production harness.
LAM None yet Moderate BORROW PATTERN ZOL->ZOE chain + work-loop partially map to LAM decompose/plan/act. Don't build LAM; steal the decomposition pipeline (doc 759's Gap 1).
MoE Router concept High USE NOW ZOL Model Gateway (cost-aware routing, fallback to SLM on budget) = MoE pattern. Route tasks: Haiku for classification, Sonnet for reasoning, Opus for escalation. Same load-balancing challenge.
VLM Pilot (computer-use) High USE NOW /meeting frame extraction (Claude vision), agent GUI control (ZOL->clicks). Budget: $2-20/image; cap at 10 images/day per agent = $5-200/day (fits $50/day cap).
SLM Ollama (llama3.1:8b) Critical USE NOW Edge fallback (Pi, cost-aware routing), inbox label classification ($0 marginal), health checks. Phi-4/Gemma-4 = better baseline for fine-tuning to ZAO voice.
MLM Bonfire embeddings Critical USE NOW Bonfire corpus search, MemoryWeaver retrieval, semantic similarity (research docs, team members). BGE-M3 for multilingual graph. Bottleneck: RLS/labeling block, not missing.
SAM None yet Low NOT RELEVANT Pixel segmentation niche. No ZAO use case in roadmap (music video editing? low priority per doc 743 WaveWarZ). Flag honestly.

Analysis: Which 3 Matter Most to ZAO, and Why?

Ranked by Impact x Imminence x Fit

1. MoE (Mixture of Experts) - IMMEDIATE

Why: ZOL Model Gateway (doc 1098) depends on it. Cost-aware model routing is the bottleneck to autonomous scale. Current state: dispatch routes to Sonnet for all tasks (expensive). With MoE thinking: route Haiku for classification/summarization, Sonnet for reasoning, Opus only for escalation + sensitive decisions. Unlocks 3-5x cost efficiency. Mimics DeepSeek-V3's sparse activation (top-K expert selection) but at model level: Haiku = cheap expert (many tasks), Sonnet = mid-expert, Opus = expert-of-last-resort. No new infrastructure needed; already have the models. Just formalize the router rules.

Concrete action: Doc 1098 finalize + add routing table (task type → model, cost budget). A/B test ZOE dispatch with/without MoE routing.

2. SLM (Small Language Model) - IMMEDIATE to Q3

Why: Edge fallback is the moat. ZAO's data (community, songs, documents) often sensitive; Ollama on Pi = zero cloud exposure. Current Ollama (llama3.1:8b) does classification OK but hallucinates on sourcing. Phi-4 (14B, synthetic data quality) would be better baseline. Also: SLMs enable on-device fine-tuning to ZAO voice (Phi-4 or Gemma-4 with ~10K examples of ZAO posts = custom model). Fallback when cloud budget exhausted (rule 5: $50/day cap). At $0.05/MTok, even 100M tokens/day = $5.

Concrete action: Test Phi-4 on Ollama (memory footprint, latency). If fits on Pi, stage as fallback. Plan fine-tuning run: 10K examples of /socials posts (Zaal, Iman, team) → Phi-4 lora adapter (ZAO voice model).

3. VLM (Vision Language Model) - Q3 to Q4

Why: Computer-use for agents is the frontier. ZOL eventually needs to screenshot a website, understand the layout, and click/type (LAM pattern). Claude Opus 4.7 vision = best-in-class for GUI control. But cost: ~$2-20/image. Budget: cap at 1 image/hour per agent = ~24/day = $50-480/day (overlaps Sonnet budget). Need to be surgical. VLMs also unlock /meeting frame extraction (keynote slides → speaker name + quote). Opportunity: SAM 3 for video (detect speaker on frame) + VLM (read speaker's name), then route to Bonfire. But that's Q4.

Concrete action: Wire Claude vision into ZOL's screenshot path (one-off for testing). Measure real cost on 100 agents x 5 screenshots/day. Decide: worth the budget trade-off vs pure text-based understanding?


What We are NOT Doing (Honestly)

  • SAM 3: Niche play. Pixel segmentation doesn't unlock revenue or autonomy for ZAO. WaveWarZ editor (doc 743) is parked; no music video pipeline yet. Flag for future, but low priority.
  • LCM (Large Concept Models): Neat research, but ecosystem immature (2026 tooling not production-ready). Multilingual is nice-to-have; English + Spanish (SONAR native) covers 90% of ZAO community. Revisit 2027 if v-LCM shipping on Claude API.
  • Direct LAM build: Don't build a LAM orchestrator (Rabbit-style). Instead, borrow the mental model (perceive → decompose → plan → act) and fold it into ZOE's workers (Gap 1 decomposition is conceptually LAM-ish already).

Appendix: 2026 Model Landscape Shifts

2025-2026 Realizations (from web research):

  1. MoE is standard, not exotic. GPT-4, Mixtral, DeepSeek-V3, Qwen all MoE. Frontier = sparse activation. Implications: assuming dense params = hallucinating. ZOL's router must think MoE-style (top-K experts).

  2. SLMs are production. Phi-4 rivals Llama-70B on reasoning. Gemma-4 does text+image+audio+video in 12B. Not toys anymore. Edge deployment in Taiwan manufacturing up 3x (2025-26). ZAO should bet on on-device fallback.

  3. LCM is the dark horse. Meta betting hard (v-LCM + v-Sonar ecosystem). Concept-space = language-agnostic reasoning. If multilingual crew grows (Berlin, Apna, AVAX), LCM is sleeper play. Monitor, don't adopt yet.

  4. VLM ≈ commodity. GPT-4o, Claude Opus 4.7, Qwen2.5-VL all competitive on MMMU, OCR, DocVQA. Price war = all <$1 per image now. Adoption barrier = just integration (add vision encoder), not capability.

  5. LAM = agent pattern. Not a specific model type, but an architecture (perception→decompose→plan→act). Rabbit's DLAM (computer controller) + OpenClaw (action toolkit) are concrete implementations. ZAO's ZOE workers already do this (Gap 1 decompose, Gap 2 dispatch). Borrow naming + formalize the pattern.


Sources & References

Research data:

ZAO internal:

  • CLAUDE.md - stack, agent fleet, Primary Surfaces
  • Memory: project_zol_farcaster_agent, project_zoe_orchestrator_locked, project_ollama_local_llm
  • Doc 759 - ZOE orchestrator gaps (reference for LAM/decomposition pattern)
  • Doc 1098 - ZOL Model Gateway (cost-aware routing, MoE pattern)
  • Doc 743 - WaveWarZ canonical positioning

Next Actions

Action Owner Due Status
Finalize ZOL Model Gateway routing table (Haiku/Sonnet/Opus by task type + cost) Zaal + engineering 2026-07-23 Blocked on doc 1098 review
Test Phi-4 on Ollama (memory, latency, quality on classification) Engineering 2026-07-30 Queued
A/B test ZOE dispatch with MoE routing (vs. current all-Sonnet) ZOE ops 2026-08-06 Plan spec needed
Wire Claude vision into ZOL screenshot path (pilot, 10 screenshots/day) Engineering 2026-08-13 Depends on budget approval
Monitor LCM ecosystem maturity (v-LCM production readiness, tooling) Research 2026-09-01 Quarterly review
Close SAM gap (document: niche, deprioritized vs. WaveWarZ roadmap) Zaal 2026-07-20 Feedback needed

Summary: One-Line Best-At for Each 8

  1. LLM: Long-form generation + multi-turn reasoning at human level
  2. LCM: Multilingual zero-shot generalization in unified concept space
  3. LAM: Multi-step computer automation (perceive → decompose → act)
  4. MoE: Cost-efficient frontier reasoning via sparse expert routing
  5. VLM: Visual reasoning + GUI control + document QA
  6. SLM: Edge deployment + on-device inference at 5-20x cost savings
  7. MLM: Semantic search + clustering via bidirectional embeddings
  8. SAM: Interactive pixel-level segmentation from prompts

Top 3 for ZAO: MoE (routing optimization) + SLM (edge fallback) + VLM (computer-use agents). Together they unblock cost-aware, resilient, autonomous scaling.