Kimi K2.0905 preview vs Kimi K2.5
Kimi K2.0905 preview (Moonshot AI, 262,144-token context) versus Kimi K2.5 (Azure AI Foundry, 262,144-token context). Kimi K2.0905 preview is cheaper by 14% on a blended token mix. Kimi K2.5 uniquely supports vision input. Across 4 public benchmarks we tracked, Kimi K2.0905 preview wins 0 and Kimi K2.5 wins 4. Use the live calculator below to plug your real usage shape into both, then route the winner via Agent Command Center for shadow A/B without code changes.
Bottom line — Kimi K2.0905 preview vs Kimi K2.5
Kimi K2.0905 preview and Kimi K2.5 target overlapping workloads but differ sharply on economics. Kimi K2.0905 preview runs roughly 14% cheaper on a blended input-plus-output token mix, which translates to approximately $300 per month at mid-market volume (100K requests/day). The gap compounds at enterprise scale, making the cost axis the first filter most teams apply when deciding between these two models.
On capability surface area, the models diverge: Kimi K2.5 supports vision input where the other does not. These differences are binary — either your workload needs the capability or it does not. Check whether any critical path in your agent pipeline depends on a capability only one model provides before committing to a migration.
Across 4 public benchmarks, Kimi K2.0905 preview leads on 0 and Kimi K2.5 leads on 4. The widest gap is on arena-elo, where Kimi K2.5 scores 119.0 points higher. Benchmarks are noisy and task-dependent — a model that leads on arena-elo may trail on code generation. The safest approach is to run both models on your own golden set before treating any benchmark as decisive.
For teams evaluating both models, the recommended path is a shadow A/B test: route production traffic through an OpenAI-compatible gateway, mirror a percentage to the candidate model, score both responses with an automated evaluator (faithfulness, tool-call correctness, latency), and compare cohort-level metrics over two weeks. Future AGI Agent Command Center supports this pattern with a single `base_url` change and built-in evaluators from the ai-evaluation SDK.
Live workload comparison
Same workload run through both models. The cheaper one is highlighted.
strategy: cost-optimized
primary:
model: kimi-k2-0905-preview
provider: moonshot
fallback:
model: kimi-k2-5
provider: azure-ai-foundry
shadow: { sample_rate: 0.05 } # mirror 5% of traffic to compare quality live| Kimi K2.0905 preview | Kimi K2.5 | |
|---|---|---|
| Input price | $0.600/M | $0.600/M |
| Output price | $2.50/M | $3.00/M |
| Context window | 262,144 | 262,144 |
| Max output | 262,144 | 262,144 |
| Function calling | ✓ | ✓ |
| Vision | — | ✓ |
| Audio input | — | — |
| Reasoning | — | — |
| Prompt caching | — | — |
| Structured output | — | — |
| Pricing verified | Jun 2, 2026 | Jun 2, 2026 |
Benchmark comparison
Side-by-side public benchmark scores. Greener bar = winner.
Cost at scale: monthly spend at three usage volumes
Estimated monthly cost assuming 1,000 input + 200 output tokens per request — a realistic chat-agent shape. Adjust your own usage in the calculator at the top of this page for an exact number.
| Scale | Kimi K2.0905 preview | Kimi K2.5 | Delta |
|---|---|---|---|
| Startup 10K requests/day | $330 /mo | $360 /mo | $30.00/mo |
| Mid-market 100K requests/day | $3,300 /mo | $3,600 /mo | $300/mo |
| Enterprise 1M requests/day | $33,000 /mo | $36,000 /mo | $3,000/mo |
At enterprise scale (1M requests/day), a difference of even ~10% in unit price compounds into thousands of dollars per month. Cached input pricing and batch tiers can shift this further — both are surfaced on each model's own page.
When to choose which
Picked from the data above — not vendor marketing. Match the rules to your workload, not the other way around.
Your inputs include screenshots, diagrams, or product photos — Kimi K2.5 accepts image input natively, the other doesn't.
On arena-elo, Kimi K2.5 scores 119.0 points higher — if your workload pattern matches that benchmark's task shape, the gap is meaningful.
Capability diff — what you gain and lose on the swap
A specific list of what each model has that the other doesn't. If your workload depends on a row in Only Kimi K2.0905 preview, switching to Kimi K2.5 means re-architecting that path (and vice versa).
- • Vision input
Capabilities both share (2)
- ✓ Function calling
- ✓ Streaming
Benchmark winners — by the numbers
For each public benchmark that has scores for both models, the higher score and the size of the gap. Benchmarks are noisy — treat anything under a 2-point delta as effectively tied.
| Benchmark | Kimi K2.0905 preview | Kimi K2.5 | Winner | Δ |
|---|---|---|---|---|
| arena-elo | 1330.0 | 1449.0 | Kimi K2.5 | +119.0 |
| livecodebench | 61.5 | 85.0 | Kimi K2.5 | +23.5 |
| mmlu-pro | 80.4 | 87.1 | Kimi K2.5 | +6.7 |
| swe-bench-verified | 65.8 | 76.8 | Kimi K2.5 | +11.0 |
Migration considerations
Concrete differences to wire through your stack before you flip traffic from one to the other.
- Kimi K2.5 has capabilities Kimi K2.0905 preview lacks: Vision input. Worth wiring through the agent design before commit.
- Provider changes from Moonshot AI to Azure AI Foundry. API authentication, rate-limit policy, regional availability, and billing all shift. Most teams route through an OpenAI-compatible gateway (e.g., Future AGI Agent Command Center) so the swap is a single `base_url` change instead of an SDK rewrite.
How to A/B test Kimi K2.0905 preview vs Kimi K2.5 in production
If you're stuck between the two, run them side-by-side on real traffic. Four steps the Future AGI team uses internally:
- 1. Point your existing OpenAI SDK at
https://gateway.futureagi.com/v1. No code change beyondbase_urland a virtual key. - 2. Mark Kimi K2.0905 preview primary, mirror 20% of traffic to Kimi K2.5 in shadow mode. Both responses are logged; only the primary is served to users.
- 3. Score every shadow response with an evaluator — faithfulness, tool-call correctness, response latency, cost. Built-in evaluators in ai-evaluation cover the common axes.
- 4. Compare cohort-level metrics after two weeks. Switch primary when the candidate wins on what matters to your workload — and stays within your latency budget.
Full walkthrough on the Agent Command Center page.
FAQ — Kimi K2.0905 preview vs Kimi K2.5
Which is cheaper, Kimi K2.0905 preview or Kimi K2.5? ▾
Kimi K2.0905 preview is cheaper by roughly 14% on a blended input + output token mix. Input prices are $0.600/M for Kimi K2.0905 preview versus $0.600/M for Kimi K2.5; output prices are $2.50/M versus $3.00/M. The exact savings depend on your input:output ratio — use the live calculator above to plug in your own request shape.
What is the context window of Kimi K2.0905 preview versus Kimi K2.5? ▾
Kimi K2.0905 preview supports up to 262,144 tokens of context. Kimi K2.5 supports up to 262,144 tokens. Kimi K2.5 has the larger window by a factor of 1.0x, which matters for long-document RAG, multi-turn agent sessions, and tasks that need to keep an entire codebase in working memory.
Do Kimi K2.0905 preview and Kimi K2.5 both support tool calling? ▾
Yes — both Kimi K2.0905 preview and Kimi K2.5 support native function calling. Both also support structured output via JSON schema, so an agent can be ported between them with the same tool definitions.
Can Kimi K2.0905 preview and Kimi K2.5 process images? ▾
Kimi K2.5 accepts native image input. Kimi K2.0905 preview does not — you would need to route image-heavy workloads through Kimi K2.5 or add a separate vision model in front of Kimi K2.0905 preview.
When should I choose Kimi K2.0905 preview over Kimi K2.5? ▾
On the data this page surfaces, Kimi K2.0905 preview is the right pick when Kimi K2.5's lower price or different capability profile aren't a fit for your workload. Run the live calculator above against your actual usage shape to confirm.
When should I choose Kimi K2.5 over Kimi K2.0905 preview? ▾
Your inputs include screenshots, diagrams, or product photos — Kimi K2.5 accepts image input natively, the other doesn't. On arena-elo, Kimi K2.5 scores 119.0 points higher — if your workload pattern matches that benchmark's task shape, the gap is meaningful.
How do I A/B test Kimi K2.0905 preview against Kimi K2.5 in production? ▾
Route both through an OpenAI-compatible gateway like Future AGI Agent Command Center with shadow mode enabled. Send 100% of traffic to your primary model, mirror 10–20% to the candidate, score every response with an evaluator (faithfulness, tool-call correctness, response time), and compare cohort-level metrics for two weeks. Switch when the candidate wins on the metrics that matter to your workload and stays within your latency budget.