Tesla FSD V13 — Occupancy + 4D World (now FSD v14)
https://www.thinkautonomous.ai/blog/occupancy-networks/
Single end-to-end neural network ingesting multi-camera video + navigation + ego-motion, maintaining an evolving 3D occupancy volume as internal world state — spatial memory persisting across frames, tracking objects that temporarily leave the camera FOV. Row originally described V13; current line is FSD (Supervised) v14 — v14.3 (2026-04-07) rewrote the AI compiler and runtime with MLIR (~20% faster reaction time) and upgraded the RL stage and vision encoder; a v14 Lite build brings V14 to HW3 cars (2026-07).
Renamed · Verified 2026-10-23 · fresh as of 2026-10-23 · 30 cited facts, 6 with a third-party source · 2 recorded changes · Suggest a correction
At a glance
- Tier
- T1
- Created
- 2024 (FSD v13 first released December 2024 to early access HW4 vehicles; wide release mid-December 2024) third-party fix
- Latest release
- FSD (Supervised) v14.3.6 for HW4 + v14 Lite for HW3 (software 2026.20.6.11, 2026-07-21) — latest found; proprietary third-party fix
- License
- not applicable — not OSS fix
- GitHub
- not applicable — no GitHub repo fix
- Pricing
- FSD (Supervised) $99/month subscription; subscription-only path on Tesla support page; requires FSD computer 3.0+ vendor fix
- Funding
- Public company (TSLA) vendor fix
- Freshness
- 2026-10-23 fix
- Layer
- agent-infra repo fix
- Licensing
- commercial vendor fix
Memory taxonomy
- storage
- vector
- retrieval
- attention
- persistence
- session
- update
- overwrite
- unit
- episode
- governance
- opaque
- conflict
- overwrite
When to use
Optimised for: real-time perception + spatial reasoning + multi-modal sensor fusion
Anti-fit: not for non-embodied / non-physical use cases
Pros & cons
Pros
Largest fleet-driven memory in any robotic system — millions of vehicles producing data daily; 4D occupancy is the memory format.
Cons
Tesla-only; closed system; safety / regulatory track record is contested.
Claims & capabilities
70k GPU-hours per training cycle. 1.5 PB fleet data from 4M+ vehicles.
Technical surface
- API surface
- searched not found
- Backend storage
- searched not found
- Deployment
- On-device in Tesla vehicles; v14.x targets HW4/AI4
- Embedding model
- searched not found
- Multi-tenancy
- searched not found
- MCP
- no MCP support advertised — vertical product, no MCP server / client integration documented
- A2A
- no A2A protocol support advertised — vertical product, no A2A integration documented
- OpenTelemetry
- no OpenTelemetry integration advertised — vendor logs/observability not publicly documented
Alternatives to Tesla FSD V13 — Occupancy + 4D World (now FSD v14) → live options ranked by shared design, licence, hosting and MCP support
Similar systems
Other vertical / domain-specific ai memory in the catalog, ranked by inbound references.
- Inworld AI T1
Formerly an NPC character engine with named Long-Term Memory; now sells realtime voice models (TTS-2, STT-1, speech-to-speech) and OpenAI-compatible open-model inference. Character/NPC memory is no longer the headline product.
- NVIDIA ReMEmbR T3
Builds long-horizon memory by captioning video segments with VILA, storing captions with timestamps + 3D position coordinates in MilvusDB. At query time, LLM iterates retrieval across text, time, and position modalities. Deployed on Nova Carter robot (Jetson Orin).
- Abridge T1
Clinician-assist ambient documentation. Source mapping: every AI-generated summary element traced back to the source utterance. Audit-and-trust layer over episodic memory. Built on proprietary 1.5M+ medical-encounter dataset.
- ASAPP GenerativeAgent T1
Treats memory as first-class architecture. Captures the digital footprint of every interaction; retrieves preference and history at engagement time. Public example: airline knowing a frequent flyer wants aisle seats with her son — preference-aware, not just history-lookup.
- BenevolentAI T1
Taken private: merged into Osaka Holdings (renamed BenevolentAI S.A.) and delisted from Euronext Amsterdam 2025-03-13. Target identification / drug repurposing / mechanism tracing. 85+ data sources, petabyte-scale, rebuilt every few weeks. Wet-lab results re-enter the graph and shift downstream predictions — institutional experimental memory closing a feedback loop. As of Oct 2026 benevolent.com leads with Qor, powered by BenevolentAI, a life-science intelligence service for biopharma decision makers and consultancies. Qor (qor.one) is a continuously updated decision-intelligence service built on BenevolentAI's knowledge graph and reasoning methods: it tracks evidence on an asset or indication (competitive position, rival readouts, comparable deals), source-verifies each claim and flags when an earlier conclusion stops holding. It targets biotech teams and life-science consultancies.
- Causaly T1
Agentic AI platform for life sciences: orchestrated scientific agents over a curated biomedical knowledge graph (plus Bio Graph and Pipeline Graph), an evidence-retrieval system and an enterprise data fabric over literature, trials and patents.
Sources (29 cited facts, 6 third-party)
Each fact links to the page it came from; the badge says who published it (how to read sources).
- What it is 3D occupancy network + temporal multi-modal fusion vendor
- What it does & what's distinct Single end-to-end neural network ingesting multi-camera video + navigation + ego-motion,… third-party
- Claims / benefits 70k GPU-hours per training cycle. 1.5 PB fleet data from 4M+ vehicles. vendor
- Modalities vision + sensor vendor
- Created 2024 (FSD v13 first released December 2024 to early access HW4 vehicles; wide release mi… third-party
- Latest release FSD (Supervised) v14.3.6 for HW4 + v14 Lite for HW3 (software 2026.20.6.11, 2026-07-21) … third-party
- Funding Public company (TSLA) vendor
- Customers / scale ~12% FSD conversion rate among Tesla installed base; HW4 customers only for v13 third-party
- Pricing FSD (Supervised) $99/month subscription; subscription-only path on Tesla support page; r… vendor
- Compliance UNECE-neutered in EU; NHTSA-regulated in US; no SOC2/ISO listed third-party
- Deployment On-device in Tesla vehicles; v14.x targets HW4/AI4 third-party
- HQ US (Austin TX) vendor
- Founders / pedigree Elon Musk (CEO/founder) vendor
- Memory volume / scale 1.5 PB fleet data from 4M+ vehicles (vendor claim) vendor
- Performance Tesla FSD V13: vendor-claimed 5-6x reduction in miles-per-disconnect from V12.x; 4D worl… vendor
- Versioning model-version semantics (policy checkpoints) vendor
- Namespace primitives robot-per-instance (per-robot memory) vendor
- Contradiction handling always overwrite (last-write-wins) vendor
- Forgetting policy episode-based memory; long-term memory module retains landmarks; episodic memory pruned … vendor
- Validated verticals robotics (humanoid, manipulation), autonomous-driving, embodied AI research vendor
- Time to running weeks-to-months (custom robot/vehicle integration) or research-only code release vendor
- Anti-fit not for non-embodied / non-physical use cases vendor
- Optimised for real-time perception + spatial reasoning + multi-modal sensor fusion vendor
- Adjacent infrastructure requires robot / vehicle hardware + sensor stack vendor
- Pros Largest fleet-driven memory in any robotic system — millions of vehicles producing data … vendor
- Cons Tesla-only; closed system; safety / regulatory track record is contested. vendor
- Project & sources occupancy-networks writeup vendor
- licensing-bucket commercial vendor
- taxonomy-three-layer-rollup agent-infra repo
What changed
Changes to this record's lifecycle, licence, release, pricing and hosting facts, from the catalog's git history. Subscribe (Atom)
- 2026-10-23
- Name “Tesla FSD V13 — Occupancy + 4D World” → “Tesla FSD V13 — Occupancy + 4D World (now FSD v14)”
- Latest release “not applicable — not OSS” → “FSD (Supervised) v14.3.6 for HW4 + v14 Lite for HW3 (software 2026.20.6.11, 2026-07-21) — latest found; proprietary” source
- 2026-10-22
- Pricing “$99/month subscription (as of Apr 2024); purchase price $8000 (discontinued Feb 2026 — subscription only)” → “FSD (Supervised) $99/month subscription; subscription-only path on Tesla support page; requires FSD computer 3.0+” source
- Deployment “HW4/AI4 Tesla vehicles only (on-device)” → “On-device in Tesla vehicles; v14.x targets HW4/AI4” source