We may earn a commission if you buy through links on this page. It never changes a score or a ranking. How we make money
Paid plans start at $5,000/mo. There's also a free plan.
Telnyx has no free tier or free trial listed. Pay as you go has a $0 monthly plan price, but inference and communications usage are billed separately. Chat completions are priced per 1M tokens, with different input, cached-input and output rates by model. Volume rate is a custom offer based on forecasted, committed usage. Enterprise costs at least $5,000 per month and adds dedicated infrastructure and IPs, private interconnect, in-region GPUs and unlimited rate limits.
01Every plan, line by line
| Plan | Billed annually | Billed monthly | Built for |
|---|---|---|---|
| Pay as you go | — | $0 | Teams starting with inference |
| Volume rate | — | Quote | Teams committing forecasted volume |
| Enterprise | — | $5,000 | Teams needing dedicated infrastructure |
Add-ons and overages
- Chat completions are billed per 1M tokens; Flex rates: DeepSeek V4.1 Flash 1017 input $0.150, cached input $0.003, output $0.600.
- Default rates per 1M tokens (input/cached/output): GLM-5.3 $1.250/$0.240/$4.000; GLM-5.3 Flash $0.135/$0.027/$0.450; DeepSeek V4.1 Flash 1017 $0.300/$0.006/$1.200; Kimi K3 $2.700/$0.270/$13.500; Kimi K3.1 (BETA) $2.000/$0.200/$10.000; Kimi K2.6 $0.665/$0.080/$4.000.
- Default rates per 1M tokens (input/cached/output): GLM-5.2 $1.000/$0.200/$4.000; MiniMax-M3 $0.270/$0.080/$1.100; Qwen 3.8 27B $0.400/$0.050/$3.000; Gemma 4 26B A4B $0.130/$0.050/$0.400; GPT OSS 120B (BETA) $0.068/$0.030/$0.300.
- Priority rates per 1M tokens (input/cached/output): GLM-5.3 Flash $1.350/$0.270/$4.500; Kimi K2.6 $1.330/$0.160/$8.000; GLM-5.2 $2.000/$0.400/$8.000; Gemma 4 31B $0.320/$0.160/$1.200.
- Decision models per 1M tokens: telnyx/decision-flash input $0.035, output included; telnyx/decision-pro input $0.200, output included.
- Embeddings (gte-large): $0.0001 per 1K tokens; speech to text: $0.003 per minute; AI-enabled storage and retrieval: $0.02 per GB per day.
- Object Storage: $0.006 per GB per month; CloudFS and SQLDB are marked custom and beta, respectively.
02What the plan grid doesn't tell you
Model and token mix
Input, cached input and output have separate rates. For example, default GLM-5.3 costs $1.250/$0.240/$4.000 per 1M tokens.
Speech and embeddings
Speech to text costs $0.003 per minute; gte-large embeddings cost $0.0001 per 1K tokens.
Storage and retrieval
AI-enabled storage and retrieval costs $0.02 per GB per day. Object Storage is $0.006 per GB per month.
Enterprise commitment
Dedicated infrastructure starts with a $5,000 monthly minimum. Volume rate requires committed volume and a custom forecast-based rate.
03Which plan should you buy?
| If you are… | Get | Why |
|---|---|---|
| Prototype or side project | Pay as you go | No monthly plan fee or contract; pay for actual usage, but there is no free tier. |
| Growing inference workload | Volume rate | A custom forecast-based rate may suit teams ready to commit volume; request a quote. |
| Team needing dedicated GPUs | Enterprise | Includes dedicated infrastructure and IPs, private interconnect, in-region GPUs and unlimited rate limits, with a $5,000 monthly minimum. |
04Is Telnyx worth the money?
Pay as you go is the practical starting point if your application needs Telnyx inference or communications and you can forecast usage costs. It has no monthly plan fee, but it is not free: model tokens, speech, embeddings and storage are metered. Volume rate is worth pricing when you can commit forecasted usage. Enterprise is for teams that need dedicated infrastructure and accept a $5,000 monthly minimum. There is no self-hosted option to avoid platform charges.
Telnyx pricing FAQ
Does Telnyx have a free tier?
No free plan or free trial is listed. Pay as you go has a $0 monthly plan price, but metered usage is billed separately. Budget for the models and services your application uses rather than treating the $0 plan price as free usage.
How are Telnyx inference models billed?
Chat completions are billed per 1M tokens, with separate rates for input, cached input and output. Rates vary by model and tier. For example, default GLM-5.3 is $1.250 per 1M input tokens, $0.240 per 1M cached input tokens and $4.000 per 1M output tokens.
What is the Telnyx Enterprise minimum?
Enterprise has a $5,000 monthly minimum spend. The listed features include dedicated infrastructure and IPs, private interconnect, in-region GPUs and unlimited rate limits. Confirm contract terms and included services with Telnyx before committing.
Does Telnyx offer annual billing?
No annual billing is shown. Pay as you go is usage-priced, Volume rate uses a custom committed-volume rate, and Enterprise has a monthly minimum. Ask Telnyx whether annual terms or discounts are available for your proposed contract.