Mistral Vibe (Remote Agents) + Mistral Medium 3.5 vs OpenAI Codex (cloud agent)
Mistral Vibe (Remote Agents) + Mistral Medium 3.5 vs OpenAI Codex (cloud agent): side-by-side comparison of two coding-agent memory systems — architecture, taxonomy, license, pricing, MCP/A2A support, and direct edges.
Mistral Vibe (Remote Agents) + Mistral Medium 3.5 · OpenAI Codex (cloud agent)
Cost & capability
| Mistral Vibe (Remote Agents) + Mistral Medium 3.5 | OpenAI Codex (cloud agent) | |
|---|---|---|
| Capability band | frontier | frontier |
| Capability composite | 76 | 78 |
| Cost tier | mid | — |
| $/Mtok input | 1.5 | — |
| $/Mtok output | 7.5 | — |
| Use cases | Code Generation Focused, Long Running Session, Scoped Agentic | Code Generation Focused, Long Running Session |
Where they differ (14)
Rows where both sides have data and the values disagree — the shortlist of dimensions that actually distinguish these two systems.
| Mistral Vibe (Remote Agents) + Mistral Medium 3.5 | OpenAI Codex (cloud agent) | |
|---|---|---|
| Capability composite | 76 | 78 |
| Use cases | Code Generation Focused, Long Running Session, Scoped Agentic | Code Generation Focused, Long Running Session |
| Type | Cloud-hosted async coding agents with teleportable session state | File-backed ~/.codex/memories + Chronicle |
| Created | 2026-05-02 (Vibe + Medium 3.5 launch) | searched not found |
| Pricing | API: $1.5/M input tokens; $7.5/M output tokens (Medium 3.5); Vibe via Pro/Team/Enterprise plans | Included in ChatGPT Plus/Pro/Business/Edu/Enterprise plans (memory feature) |
| Funding | $2.71B total raised; Series C €1.7B (ASML lead) Sep 2025; $830M debt Mar 2026 | OpenAI $57B+ raised; $300B+ valuation |
| Backend storage | custom (Mistral-managed sandboxes) | custom |
| Deployment | Managed cloud (la Plateforme + Le Chat) + NVIDIA NIM containers | Both (cloud agent via ChatGPT platform + CLI tool local) |
| API surface | REST + CLI (Mistral CLI) + Le Chat web; SDK Python/JS | REST (OpenAI Responses API; Codex CLI wraps it) |
| Embedding | Mistral embed-v2 (default; multiple supported) | locked |
| Multi-tenancy | hard-isolation (per-session sandbox) | hard-isolation |
| MCP | native (first-party — Mistral Agents support MCP) | native (first-party) |
| Optimised for | async parallel coding agents with teleportable session state; SWE-Bench-leading model (77.6%); EU-jurisdiction stack | file-backed memory under ~/.codex/memories |
| Anti-fit | not for non-Mistral stacks; not for users requiring full transparency of memory eviction policy; not for offline/air-gapped use | not for non-coding tasks |
At a glance
| Mistral Vibe (Remote Agents) + Mistral Medium 3.5 | OpenAI Codex (cloud agent) | |
|---|---|---|
| Section | Coding-agent memory | Coding-agent memory |
| Tier | T1 | T1 |
| Type | Cloud-hosted async coding agents with teleportable session state | File-backed ~/.codex/memories + Chronicle |
| Created | 2026-05-02 (Vibe + Medium 3.5 launch) | searched not found |
| Latest release | Medium 3.5 weights on Hugging Face (modified MIT); Vibe service closed | — |
| License | Modified MIT (Medium 3.5 weights); Vibe service closed | — |
| Pricing | API: $1.5/M input tokens; $7.5/M output tokens (Medium 3.5); Vibe via Pro/Team/Enterprise plans | Included in ChatGPT Plus/Pro/Business/Edu/Enterprise plans (memory feature) |
| Funding | $2.71B total raised; Series C €1.7B (ASML lead) Sep 2025; $830M debt Mar 2026 | OpenAI $57B+ raised; $300B+ valuation |
| Backend storage | custom (Mistral-managed sandboxes) | custom |
| Deployment | Managed cloud (la Plateforme + Le Chat) + NVIDIA NIM containers | Both (cloud agent via ChatGPT platform + CLI tool local) |
| API surface | REST + CLI (Mistral CLI) + Le Chat web; SDK Python/JS | REST (OpenAI Responses API; Codex CLI wraps it) |
| Embedding | Mistral embed-v2 (default; multiple supported) | locked |
| Multi-tenancy | hard-isolation (per-session sandbox) | hard-isolation |
| MCP | native (first-party — Mistral Agents support MCP) | native (first-party) |
| A2A | searched not found | searched not found |
| OpenTelemetry | searched not found | searched not found |
| Optimised for | async parallel coding agents with teleportable session state; SWE-Bench-leading model (77.6%); EU-jurisdiction stack | file-backed memory under ~/.codex/memories |
| Anti-fit | not for non-Mistral stacks; not for users requiring full transparency of memory eviction policy; not for offline/air-gapped use | not for non-coding tasks |
Taxonomy
| Axis | Mistral Vibe (Remote Agents) + Mistral Medium 3.5 | OpenAI Codex (cloud agent) |
|---|---|---|
| storage | kv | file |
| retrieval | injection | injection |
| persistence | cross-session | long-term |
| update | append-only | consolidation |
| unit | episode | summary |
| governance | opaque | inspectable |
| conflict | none | none |
Pros & cons
Mistral Vibe (Remote Agents) + Mistral Medium 3.5
Pros: Async parallel coding agents with teleportable session state; broad ecosystem integrations (GitHub/Linear/Jira/Sentry); SWE-Bench-leading model; weights open; EU-jurisdiction stack.
Cons: Vibe service closed; SWE-Bench claim is vendor self-report; no third-party reproduction yet; pricing tied to Pro/Team/Enterprise tiers; opaque memory eviction policy.
OpenAI Codex (cloud agent)
Pros: Tight integration with the OpenAI model + tool-use roadmap; benefits directly from frontier model improvements.
Cons: Memory is shallow — primarily session-scoped; no long-running project memory yet.