API cost decision in 10 seconds

NewTernary Bonsai 2 9B vs Step 5 Preview

Pick Ternary Bonsai 2 9B for lower cost; pick Step 5 Preview only if the larger context window matters more.

Page updated: 2026-10-27T20:08:11Z Data confirmed: 2026-10-28T05:08:04Z Prices normalized to USD per 1M tokens Sample workload: 1M input + 500K output

Budget verdict

Pick Ternary Bonsai 2 9B for lower cost; pick Step 5 Preview only if the larger context window matters more.

On the standard 1M input plus 500K output workload, Ternary Bonsai 2 9B is estimated at $0.15 vs $2.35 for Step 5 Preview, saving $2.21 (93.8% lower).

Cost-first pickTernary Bonsai 2 9B

Context-first pickStep 5 Preview

Sample savings$2.2193.8%

10x traffic gap$22.05

Step 5 Preview has more context, but Ternary Bonsai 2 9B saves $2.21 on the standard workload. At 10x that traffic, the same price gap is about $22.05. Use the calculator below to replace the sample workload with your own token volume.

Cost sensitivity

Workload Sensitivity

Same prices, different token mixes.

Ternary Bonsai 2 9B stays cheaper across input-heavy, balanced, and output-heavy sample workloads.

Workload shape	Token mix	Better pick	Ternary Bonsai 2 9B	Step 5 Preview
Input-heavy / RAG	5M input + 500K output	Ternary Bonsai 2 9B	$0.47	$6.35
Balanced workload	1M input + 1M output	Ternary Bonsai 2 9B	$0.21	$3.7
Output-heavy chatbot	1M input + 5M output	Ternary Bonsai 2 9B	$0.73	$14.5

Cheaper input Ternary Bonsai 2 9B $0.08 vs $1 / 1M

Ternary Bonsai 2 9B is $0.92 cheaper per 1M input tokens (92% lower; 12.5x difference).

Cheaper output Ternary Bonsai 2 9B $0.13 vs $2.7 / 1M

Ternary Bonsai 2 9B is $2.57 cheaper per 1M output tokens (95.2% lower; 20.8x difference).

Larger context Step 5 Preview 262.14K vs 1.05M

Step 5 Preview has 786.43K more context (4x larger).

Sample workload Ternary Bonsai 2 9B $0.15 vs $2.35

Ternary Bonsai 2 9B is $2.21 cheaper on the standard workload (93.8% lower).

Estimate your workload cost

Your Workload Cost

Prices are normalized to USD per 1M tokens.

Monthly input tokens Monthly output tokens

Ternary Bonsai 2 9B Calculating… Estimated API cost

Step 5 Preview Calculating… Estimated API cost

Cheaper for this workload Calculating… Difference: calculating…

This estimate uses normalized public API pricing per 1M tokens. It is a planning aid, not a billing quote. Verify provider pricing, limits, and terms before production use.

Quick Decision

Verdict

Ternary Bonsai 2 9B has the lower input price; Ternary Bonsai 2 9B has the lower output price; Step 5 Preview offers the larger context window. For the 1M input plus 500K output sample, Ternary Bonsai 2 9B is cheaper for the standard workload.

For a 1M input token plus 500K output token workload, the estimated API cost is $0.15 for Ternary Bonsai 2 9B and $2.35 for Step 5 Preview.

Best Fit

Choose Ternary Bonsai 2 9B when you care most about lower input-token price, and lower output-token price.

Choose Step 5 Preview when you care most about larger context window.

Decision Notes

On the standard 1M input plus 500K output workload, Ternary Bonsai 2 9B is estimated at $0.15 vs $2.35 for Step 5 Preview, saving $2.21 (93.8% lower).
Ternary Bonsai 2 9B is $2.21 cheaper on the standard workload (93.8% lower).
Ternary Bonsai 2 9B is $0.92 cheaper per 1M input tokens (92% lower; 12.5x difference).
Ternary Bonsai 2 9B is $2.57 cheaper per 1M output tokens (95.2% lower; 20.8x difference).
Step 5 Preview has 786.43K more context (4x larger).

Head-to-Head Specs

Feature	NewTernary Bonsai 2 9B (PrismML)	Step 5 Preview (StepFun)
Input Price prompt tokens per 1M	$0.08	$1
Completion Price per 1M tokens	$0.13	$2.7
Sample Workload Cost 1M input + 500K output	$0.15	$2.35
Context Window	262.14K	1.05M
Release Date	2026-10-16	2026-10-08

Use-Case Decision Matrix

Use case	Better pick	Why
Budget-constrained production	Ternary Bonsai 2 9B	On the standard 1M input plus 500K output workload, Ternary Bonsai 2 9B is estimated at $0.15 vs $2.35 for Step 5 Preview, saving $2.21 (93.8% lower).
High-volume input processing	Ternary Bonsai 2 9B	Lower prompt-token price matters most when prompts, retrieved passages, or documents dominate the bill.
Long responses and chatbots	Ternary Bonsai 2 9B	Lower output-token price matters most when assistants generate many completion tokens.
RAG or long-document work	Step 5 Preview	A larger context window leaves more room for retrieved passages, conversation history, or source files.

Related Alternatives

Same-provider lower-cost swaps

Step 3.5 Flash can replace Step 5 Preview when lower sample workload cost matters most: $0.25.
Step 3.7 Flash can replace Step 5 Preview when lower sample workload cost matters most: $0.77.

Larger context near this budget

Grok 4.20 Multi-Agent offers 2M context with $2.5 sample workload cost.
Grok 4.20 offers 2M context with $2.5 sample workload cost.
Llama 4 Scout offers 1.31M context with $0.25 sample workload cost.
GPT-6 Luna offers 1.05M context with $0.35 sample workload cost.

Popular competitors

Kimi K3.1 Pro · MoonshotAI · #1
DeepSeek V4.1 Flash 1017 · DeepSeek · #2
DeepSeek V4.1 Flash · DeepSeek · #3
GLM 5.3 Flash · Z.ai · #4

Cheaper alternatives

Review low-cost models sorted by a standard 1M input plus 500K output workload.

Open cheapest models

Larger context alternatives

Find models with larger context windows for RAG, long documents, and codebase review.

Open largest context models

Provider catalogs

Compare models within provider hubs before choosing a final API vendor.

Open provider hubs

PrismML catalog

Review all tracked PrismML models before deciding whether this matchup is the right shortlist.

Open PrismML models

StepFun catalog

Check other StepFun models with comparable pricing, context, or release timing.

Open StepFun models

Ternary Bonsai 2 9B

Ternary Bonsai 2 9B is PrismML’s compact, open-weight language model for conversational AI, reasoning, and tool use. Built on Qwen3.5, it uses ternary weights packed into 2 bits for efficient...

Step 5 Preview

Step 5 Preview is StepFun's flagship model for agentic work, built on a sparse Mixture-of-Experts architecture (27B active / 600B total parameters). It performs strongly in software engineering and professional...